Hugging Face's logo

Datasets:
shubhamg2208
/
lexicap

Tasks:
Text Classification
Text Generation
Sub-tasks:
sentiment-analysis
dialogue-modeling
language-modeling
Languages:
English
Size:
n<1K
Tags:
karpathy,whisper,openai
Dataset card Files Community
lexicap / vtt /episode_007_small.vtt
Shubham Gupta
Add readme and files
a3be5d0
raw
history blame
160 kB
WEBVTT
00:00.000 --> 00:02.400
The following is a conversation with Jeff Atwood.
00:02.400 --> 00:05.840
He is the cofounder of Stack Overflow and Stack Exchange,
00:05.840 --> 00:08.240
websites that are visited by millions of people
00:08.240 --> 00:09.560
every single day.
00:09.560 --> 00:13.880
Much like with Wikipedia, it is difficult to understate
00:13.880 --> 00:16.680
the impact on global knowledge and productivity
00:16.680 --> 00:19.520
that these networks of sites have created.
00:19.520 --> 00:24.000
Jeff is also the author of the famed blog, Coding Horror,
00:24.000 --> 00:27.640
and the founder of Discourse, an open source software project
00:27.640 --> 00:31.600
that seeks to improve the quality of our online community
00:31.600 --> 00:33.440
discussions.
00:33.440 --> 00:36.640
This conversation is part of the MIT course
00:36.640 --> 00:38.480
on artificial journal intelligence
00:38.480 --> 00:40.840
and the artificial intelligence podcast.
00:40.840 --> 00:43.680
If you enjoy it, subscribe on YouTube, iTunes,
00:43.680 --> 00:45.640
or your podcast provider of choice,
00:45.640 --> 00:47.480
or simply connect with me on Twitter
00:47.480 --> 00:51.000
at Lex Friedman, spelled FRID.
00:51.000 --> 00:56.200
And now, here's my conversation with Jeff Atwood.
00:56.200 --> 00:59.560
Having cocreated and managed for a few years the world's
00:59.560 --> 01:02.960
largest community of programmers in Stack Overflow
01:02.960 --> 01:08.600
10 years ago, what do you think motivates most programmers?
01:08.600 --> 01:13.320
Is it fame, fortune, glory, process of programming itself,
01:13.320 --> 01:16.360
or is it the sense of belonging to a community?
01:16.360 --> 01:18.000
It's puzzles, really.
01:18.000 --> 01:22.280
I think it's this idea of working on puzzles
01:22.280 --> 01:25.800
independently of other people and just solving a problem,
01:25.800 --> 01:29.720
sort of on your own, almost, although nobody really
01:29.720 --> 01:32.240
works alone in programming anymore.
01:32.240 --> 01:36.520
But I will say there's an aspect of hiding yourself away
01:36.520 --> 01:39.600
and just beating on a problem until you solve it.
01:39.600 --> 01:41.040
Like, brute force, basically, to me,
01:41.040 --> 01:42.200
is what a lot of programming is.
01:42.200 --> 01:45.040
It's like, the computer's so fast that you can do things
01:45.040 --> 01:46.560
that would take forever for a human,
01:46.560 --> 01:49.240
but you can just do them so many times and so often
01:49.240 --> 01:52.000
that you get the answer.
01:52.000 --> 01:55.520
You're saying just the pure act of tinkering with the code
01:55.520 --> 01:59.600
is the thing that drives most probably the joy, the struggle
01:59.600 --> 02:03.880
balance within the joy of overcoming the brute force
02:03.880 --> 02:06.760
process of pain and suffering that eventually leads
02:06.760 --> 02:09.040
to something that actually works?
02:09.040 --> 02:10.080
Well, data's fun, too.
02:10.080 --> 02:12.240
Like, there's this thing called the shuffling problem.
02:12.240 --> 02:14.960
Like, the naive shuffle that most programmers write
02:14.960 --> 02:15.800
has a huge flaw.
02:15.800 --> 02:17.520
And there's a lot of articles online about this
02:17.520 --> 02:20.040
because it can be really bad if you're like a casino
02:20.040 --> 02:22.240
and you have an unsophisticated programmer writing
02:22.240 --> 02:23.480
your shuffle algorithm.
02:23.480 --> 02:25.120
There's surprising ways to get this wrong.
02:25.120 --> 02:27.240
But the neat thing is, the way to figure that out
02:27.240 --> 02:29.040
is just to run your shuffle a bunch of times
02:29.040 --> 02:32.200
and see how many orientations of cards you get.
02:32.200 --> 02:34.680
You should get an equal distribution of all the cards.
02:34.680 --> 02:36.480
And with the naive method of shuffling,
02:36.480 --> 02:38.400
if you just look at the data, if you just brute force it
02:38.400 --> 02:40.880
and say, OK, I don't know what's going to happen.
02:40.880 --> 02:43.320
You just write a program that does it a billion times
02:43.320 --> 02:46.080
and then see what the buckets look like of the data.
02:46.080 --> 02:48.200
And the money haul problem is another example of that,
02:48.200 --> 02:50.640
where you have three doors and somebody gives you
02:50.640 --> 02:52.720
information about another door.
02:52.720 --> 02:55.000
So the correct answer is you should always switch.
02:55.000 --> 02:56.720
The money haul problem, which is not intuitive.
02:56.720 --> 02:58.080
It freaks people out all the time.
02:58.080 --> 02:59.800
But you can solve it with data.
02:59.800 --> 03:04.400
If you write a program that does the money haul game
03:04.400 --> 03:06.880
and then never switches, then always switches, just compare,
03:06.880 --> 03:09.680
you would immediately see that you don't have to be smart.
03:09.680 --> 03:11.600
You don't have to figure out the answer algorithmically.
03:11.600 --> 03:13.000
You can just brute force it out with data
03:13.000 --> 03:14.400
and say, well, I know the answer is this
03:14.400 --> 03:16.160
because I ran the program a billion times
03:16.160 --> 03:18.960
and these are the data buckets that I got from it.
03:18.960 --> 03:20.280
So empirically, you find it.
03:20.280 --> 03:22.320
But what's the joy of that?
03:22.320 --> 03:26.320
So for you, for you personally, outside of family,
03:26.320 --> 03:29.440
what motivates you in this process?
03:29.440 --> 03:31.000
Well, to be honest, I don't really
03:31.000 --> 03:32.520
write a lot of code anymore.
03:32.520 --> 03:35.240
What I do at discourse is like, managerry stuff, which
03:35.240 --> 03:38.240
I always kind of despise as a programmer.
03:38.240 --> 03:40.000
You think of managers as people who don't really
03:40.000 --> 03:42.320
do anything themselves.
03:42.320 --> 03:44.000
But the weird thing about code is you
03:44.000 --> 03:45.880
realize that language is code.
03:45.880 --> 03:48.320
The ability to direct other people
03:48.320 --> 03:51.480
lets you get more stuff done than you could by yourself
03:51.480 --> 03:52.280
anyway.
03:52.280 --> 03:53.680
You said language is code?
03:53.680 --> 03:54.200
Language is code.
03:54.200 --> 03:56.040
Meaning communication with other humans?
03:56.040 --> 03:56.640
Yes, it is.
03:56.640 --> 03:58.480
You can think of it as a systematic.
03:58.480 --> 04:01.320
So what does it like to be?
04:01.320 --> 04:03.160
What makes, before we get into programming,
04:03.160 --> 04:04.320
what makes a good manager?
04:04.320 --> 04:05.800
What makes a good leader?
04:05.800 --> 04:08.320
Well, I think a leader, it's all about leading by example,
04:08.320 --> 04:11.440
first of all, like sort of doing and being the things
04:11.440 --> 04:12.480
that you want to be.
04:12.480 --> 04:13.640
Now, this can be kind of exhausting.
04:13.640 --> 04:15.120
Particularly if you have kids because you realize
04:15.120 --> 04:17.600
that your kids are watching you all the time,
04:17.600 --> 04:20.600
like even in ways that you've stopped seeing yourself.
04:20.600 --> 04:22.160
Like the hardest person to see on the planet
04:22.160 --> 04:23.640
is really yourself.
04:23.640 --> 04:25.960
It's a lot of you see other people and make judgments
04:25.960 --> 04:28.520
about them, but yourself, like you're super biased.
04:28.520 --> 04:30.840
You don't actually see yourself the way other people see you.
04:30.840 --> 04:33.040
Often, you're very, very hard on yourself in a way
04:33.040 --> 04:34.760
that other people really aren't going to be.
04:34.760 --> 04:37.880
So that's one of the insights is you've
04:37.880 --> 04:39.560
got to be really diligent about thinking,
04:39.560 --> 04:41.480
like, am I behaving in a way that
04:41.480 --> 04:44.720
represents how I want other people to behave?
04:44.720 --> 04:46.120
Like leading through example.
04:46.120 --> 04:47.960
There's a lot of examples of leaders
04:47.960 --> 04:49.320
that really mess this up.
04:49.320 --> 04:52.080
Like they make decisions that are like, wow, that's why.
04:52.080 --> 04:55.000
It's just it's a bad example for other people.
04:55.000 --> 04:57.440
So I think leading by example is one.
04:57.440 --> 04:59.680
The other one, I believe it is working really hard.
04:59.680 --> 05:01.560
I don't mean like working exhaustively,
05:01.560 --> 05:06.760
but like showing a real passion for the problem.
05:06.760 --> 05:09.040
Like not necessarily your solution to the problem,
05:09.040 --> 05:11.240
but the problem itself is just one that you really believe in.
05:11.240 --> 05:13.120
With discourse, for example, the problem
05:13.120 --> 05:15.040
that we're looking at, which is my current project,
05:15.040 --> 05:17.920
is how do you get people in groups
05:17.920 --> 05:20.080
to communicate in a way that doesn't like break down
05:20.080 --> 05:21.800
into the howling of wolves, right?
05:21.800 --> 05:23.600
Like how do you deal with trolling?
05:23.600 --> 05:25.080
Not like technical problems of how
05:25.080 --> 05:26.440
do I get people to post paragraphs?
05:26.440 --> 05:27.520
How do I get people to use bold?
05:27.520 --> 05:29.520
How do I get people to use complete sentences?
05:29.520 --> 05:30.920
Although those are problems as well.
05:30.920 --> 05:33.560
But how do I get people to get along with each other, right?
05:33.560 --> 05:35.920
And then solve whatever problem it is they set out to solve,
05:35.920 --> 05:38.280
or reach some consensus on discussion,
05:38.280 --> 05:40.080
or just not hurt each other, even, right?
05:40.080 --> 05:41.960
Like maybe it's a discussion that doesn't really matter,
05:41.960 --> 05:43.600
but are people yelling at each other, right?
05:43.600 --> 05:44.280
And why, right?
05:44.280 --> 05:46.760
Like that's not the purpose of this kind of communication.
05:46.760 --> 05:52.960
So I would say leadership is about setting an example,
05:52.960 --> 05:54.920
doing the things that represent what you want to be,
05:54.920 --> 05:57.000
and making sure that you're actually doing those things.
05:57.000 --> 05:59.360
And there's a trick to that too, because the things you don't
05:59.360 --> 06:02.400
do also say a lot about what you are.
06:02.400 --> 06:03.960
Yeah, so let's pause on that one.
06:03.960 --> 06:05.360
So those two things are fascinating.
06:05.360 --> 06:08.080
So how do you have as a leader that self awareness?
06:08.080 --> 06:10.280
So you just said it's really hard to be self aware.
06:10.280 --> 06:12.920
So for you personally, or maybe for other leaders
06:12.920 --> 06:14.760
you've seen or look up to, how do you
06:14.760 --> 06:18.680
know both that the things you're doing
06:18.680 --> 06:20.920
are the wrong things to be doing, the way you speak
06:20.920 --> 06:23.640
to others, the way you behave, and the things you're not doing?
06:23.640 --> 06:25.200
How do you get that signal?
06:25.200 --> 06:26.440
There's two aspects to that.
06:26.440 --> 06:28.760
One is like processing feedback that you're getting.
06:28.760 --> 06:30.240
So how do you get feedback?
06:30.240 --> 06:30.640
Well, right.
06:30.640 --> 06:32.280
So are you getting feedback, right?
06:32.280 --> 06:33.480
Like so one way we do it, for example,
06:33.480 --> 06:34.840
at discourse we have three cofounders,
06:34.840 --> 06:38.160
and we periodically talk about decisions before we make them.
06:38.160 --> 06:41.080
So it's not like one person can make a mistake, or like,
06:41.080 --> 06:43.240
wow, they're going to be misunderstanding things.
06:43.240 --> 06:45.560
So it's part of group consensus of leadership
06:45.560 --> 06:47.800
is like it's good to have, I think, systems
06:47.800 --> 06:50.320
where there's one leader, and that leader has the rule
06:50.320 --> 06:53.240
of absolute law, are just really dangerous in my experience.
06:53.240 --> 06:55.400
For communities, for example, if you have a community that's
06:55.400 --> 06:57.560
run by one person, that one person makes all decisions,
06:57.560 --> 06:59.400
that person's going to have a bad day.
06:59.400 --> 07:02.360
Something can happen to that person.
07:02.360 --> 07:03.280
There's a lot of variables.
07:03.280 --> 07:05.080
So first, when you think about leadership,
07:05.080 --> 07:07.080
have multiple people doing leadership,
07:07.080 --> 07:08.920
and have them talk amongst each other.
07:08.920 --> 07:11.360
So you're giving each other feedback about the decisions
07:11.360 --> 07:12.200
that they're making.
07:12.200 --> 07:13.720
And then when you do get feedback,
07:13.720 --> 07:16.320
I think there's that little voice in your head, right?
07:16.320 --> 07:19.120
Or your gut, or wherever you want to put it in your body.
07:19.120 --> 07:22.600
I think that voice is really important.
07:22.600 --> 07:25.200
I think most people who have any kind of moral compass
07:25.200 --> 07:27.360
or want to do, most people want to do the right thing.
07:27.360 --> 07:28.440
I do believe that.
07:28.440 --> 07:30.920
I mean, there might be a handful of sociopaths out there
07:30.920 --> 07:32.760
that don't, but most people, they
07:32.760 --> 07:34.960
want other people to think of them as a good person.
07:34.960 --> 07:35.720
Why wouldn't you, right?
07:35.720 --> 07:36.920
Do you want people to despise you?
07:36.920 --> 07:38.120
I mean, that's just weird, right?
07:38.120 --> 07:40.520
So you have that little voice that the angel and devil
07:40.520 --> 07:43.040
on your shoulder talking to you about what you're doing,
07:43.040 --> 07:44.600
how you're doing, how does it make you feel
07:44.600 --> 07:46.320
to make these decisions, right?
07:46.320 --> 07:49.280
And I think having some attunement to that voice
07:49.280 --> 07:50.240
is important.
07:50.240 --> 07:52.400
But you said that voice also for,
07:52.400 --> 07:55.160
I think it's a programmer situation, too,
07:55.160 --> 07:57.120
where sometimes the devil on the shoulder
07:57.120 --> 07:59.800
is a little too loud.
07:59.800 --> 08:03.480
So you're a little too self critical for a lot of developers,
08:03.480 --> 08:05.600
especially when you have introverted personality.
08:05.600 --> 08:07.600
How do you struggle with the self criticism
08:07.600 --> 08:10.680
or the criticism of others?
08:10.680 --> 08:11.920
One of the things of leadership is
08:11.920 --> 08:14.480
to do something that's potentially unpopular
08:14.480 --> 08:16.240
or what people doubt you and you still
08:16.240 --> 08:18.320
go through with the decision.
08:18.320 --> 08:20.720
So what's that balance like?
08:20.720 --> 08:22.760
I think you have to walk people through your decision
08:22.760 --> 08:23.640
making, right?
08:23.640 --> 08:25.600
This is where blogging is really important in communication.
08:25.600 --> 08:26.080
It's so important.
08:26.080 --> 08:27.960
Again, code language is just another kind of code.
08:27.960 --> 08:30.240
It's like, here is the program by which
08:30.240 --> 08:32.960
I arrived at the conclusion that I'm going to reach, right?
08:32.960 --> 08:34.720
It's one thing to say, this is a decision.
08:34.720 --> 08:35.680
It's final.
08:35.680 --> 08:36.760
Deal with it, right?
08:36.760 --> 08:38.760
That's not usually satisfying to people.
08:38.760 --> 08:40.360
But if you say, look, we've been thinking
08:40.360 --> 08:41.360
on this problem for a while.
08:41.360 --> 08:42.720
Here's some stuff that's happened.
08:42.720 --> 08:44.080
Here's what we think is right.
08:44.080 --> 08:44.920
Here's our goals.
08:44.920 --> 08:46.120
Here's what we want to achieve.
08:46.120 --> 08:47.920
And we've looked at these options.
08:47.920 --> 08:50.560
And we think this of the available options is the best option.
08:50.560 --> 08:51.800
People will be like, oh, OK.
08:51.800 --> 08:53.080
Maybe I don't totally agree with you,
08:53.080 --> 08:54.840
but I can kind of see where you're coming from.
08:54.840 --> 08:57.320
And I see it's not just arbitrary decision delivered
08:57.320 --> 09:00.240
from a cloud of flames in the sky.
09:00.240 --> 09:03.400
It's like a human trying to reach some kind of consensus
09:03.400 --> 09:04.320
about goals.
09:04.320 --> 09:05.760
And their goals might be different than yours.
09:05.760 --> 09:06.840
That's completely legit, right?
09:06.840 --> 09:09.200
But if you're making that clear, it's like, oh, well,
09:09.200 --> 09:11.080
the reason we don't agree is because we have totally
09:11.080 --> 09:11.840
different goals, right?
09:11.840 --> 09:12.840
How could we agree?
09:12.840 --> 09:14.240
It's not that you're a bad person.
09:14.240 --> 09:16.640
It's that we have radically different goals in mind
09:16.640 --> 09:18.240
when we started looking at this problem.
09:18.240 --> 09:22.040
And the other one you said is passion or hard work, sorry.
09:22.040 --> 09:23.920
Well, those are tied together in my mind.
09:23.920 --> 09:25.240
Let's say hard work and passion.
09:25.240 --> 09:27.520
For me, I just really love the problem discourse
09:27.520 --> 09:29.720
is setting out to solve because, in a way,
09:29.720 --> 09:33.000
it's like there's a vision of the world where it all devolves
09:33.000 --> 09:35.160
into Facebook basically owning everything
09:35.160 --> 09:37.040
in every aspect of human communication, right?
09:37.040 --> 09:39.920
And this has always been kind of a scary world for me.
09:39.920 --> 09:42.400
First, because I think Facebook is really good at execution.
09:42.400 --> 09:43.280
I got to compliment them.
09:43.280 --> 09:45.400
They're very competent in terms of what they're doing.
09:45.400 --> 09:48.160
But Facebook has not much of a moral compass
09:48.160 --> 09:51.040
in terms of Facebook cares about Facebook, really.
09:51.040 --> 09:53.560
They don't really care about you and your problems.
09:53.560 --> 09:56.520
What they care about is how big they can make Facebook, right?
09:56.520 --> 09:58.760
Is that you talking about the company or just the mechanism
09:58.760 --> 09:59.960
of how Facebook works?
09:59.960 --> 10:01.320
Kind of both, really, right?
10:01.320 --> 10:03.640
And the idea with discourse, the reason I'm so passionate
10:03.640 --> 10:05.600
about it is because I believe every community should have
10:05.600 --> 10:07.120
the right to own themselves, right?
10:07.120 --> 10:09.680
Like they should have their own software that they can run
10:09.680 --> 10:11.080
that belongs to them.
10:11.080 --> 10:12.920
That's their space where they can set the rules.
10:12.920 --> 10:15.200
And if they don't like it, they can move to different hosting
10:15.200 --> 10:18.120
or whatever they need to happen can happen.
10:18.120 --> 10:21.880
But this idea of a company town where all human communication
10:21.880 --> 10:25.000
is implicitly owned by WhatsApp, Instagram, and Facebook.
10:25.000 --> 10:27.120
And it's really disturbing, too, because Facebook
10:27.120 --> 10:27.840
is really smart.
10:27.840 --> 10:29.360
Like I said, they're great at execution.
10:29.360 --> 10:30.720
Buying in WhatsApp and buying Instagram
10:30.720 --> 10:33.600
were incredibly smart decisions.
10:33.600 --> 10:35.080
And they also do this thing on, if you know,
10:35.080 --> 10:37.480
but they have this VPN software that they give away
10:37.480 --> 10:38.840
for free on smartphones.
10:38.840 --> 10:42.000
And it indirectly feeds all the data about the traffic
10:42.000 --> 10:42.880
back to Facebook.
10:42.880 --> 10:44.640
So they can see what's actually getting popular
10:44.640 --> 10:46.000
through the VPNs, right?
10:46.000 --> 10:48.400
They have low level access to the network data
10:48.400 --> 10:50.480
because users have let them have that.
10:50.480 --> 10:54.160
So let's take a small pause here.
10:54.160 --> 10:56.760
First of all, discourse, can you talk about,
10:56.760 --> 11:00.920
can you lay out the land of all the different ways
11:00.920 --> 11:01.880
you can have communities?
11:01.880 --> 11:03.880
So there's Stack Overflow that you've built.
11:03.880 --> 11:05.320
There's discourse.
11:05.320 --> 11:07.320
So Stack Overflow is kind of like a Wiki.
11:07.320 --> 11:09.240
Wikipedia you talk about.
11:09.240 --> 11:12.000
And it's a very specific scalpel, very focused.
11:12.000 --> 11:13.560
So what is the purpose of discourse?
11:13.560 --> 11:16.120
And maybe contrast that with Facebook.
11:16.120 --> 11:18.120
First of all, say what is discourse?
11:18.120 --> 11:18.960
Yeah.
11:18.960 --> 11:19.800
Start from the beginning.
11:19.800 --> 11:20.620
Well, let me start from the very beginning.
11:20.620 --> 11:22.040
So Stack Overflow is very structured,
11:22.040 --> 11:24.720
Wiki style Q&A for programmers, right?
11:24.720 --> 11:26.800
And that was the problem we first worked on.
11:26.800 --> 11:29.080
And when we started, we thought it was discussions
11:29.080 --> 11:32.080
because we looked at like programming forums and other things,
11:32.080 --> 11:33.520
but we quickly realized we were doing Q&A,
11:33.520 --> 11:37.040
which is a very narrow subset of human communication, right?
11:37.040 --> 11:38.920
So when you started Stack Overflow,
11:38.920 --> 11:41.920
you thought you didn't even know the Q&A.
11:41.920 --> 11:43.400
You didn't know it would be Q&A.
11:43.400 --> 11:44.320
Well, we didn't know.
11:44.320 --> 11:45.280
We had an idea of like, okay,
11:45.280 --> 11:47.000
these are things that we see working online.
11:47.000 --> 11:47.840
We had a goal, right?
11:47.840 --> 11:50.120
Our goal was there was this site,
11:50.120 --> 11:52.680
Experts Exchange with a very unfortunate name.
11:52.680 --> 11:53.960
Thank you for killing that site.
11:53.960 --> 11:54.840
Yeah, I know, right?
11:54.840 --> 11:56.520
Like a lot of people don't remember it anymore,
11:56.520 --> 11:57.360
which is great.
11:57.360 --> 11:58.200
Like that's the measure of success.
11:58.200 --> 11:59.360
If people don't remember the thing
11:59.360 --> 12:02.680
that you were trying to replace, then you've totally won.
12:02.680 --> 12:05.680
So it was a place to get answers to programming questions,
12:05.680 --> 12:07.520
but it wasn't clear if it was like focused Q&A,
12:07.520 --> 12:08.840
if it was a discussion,
12:08.840 --> 12:10.400
there were plenty of programming forums.
12:10.400 --> 12:11.400
So we weren't really sure.
12:11.400 --> 12:13.520
We were like, okay, we'll take aspects of dig and Reddit,
12:13.520 --> 12:15.800
like voting were very important,
12:15.800 --> 12:17.480
reordering answers based on votes,
12:17.480 --> 12:19.640
Wiki style stuff of like being able to edit posts,
12:19.640 --> 12:21.320
not just your posts, but other people's posts
12:21.320 --> 12:23.920
to make them better and keep them more up to date.
12:23.920 --> 12:26.240
Ownership of blogging of like, okay, this is me.
12:26.240 --> 12:28.120
I'm saying this in my voice, you know,
12:28.120 --> 12:29.840
this is the stuff that I know.
12:29.840 --> 12:33.360
And you know, you get your reputation accrues to you
12:33.360 --> 12:34.880
and it's pure recognition.
12:34.880 --> 12:37.080
So you asked earlier, like what motivates programmers?
12:37.080 --> 12:40.080
I think pure recognition motivates them a lot.
12:40.080 --> 12:41.880
That was one of the key insights of Stack Overflow
12:41.880 --> 12:43.720
was like recognition from your peers
12:43.720 --> 12:44.960
is why things get done.
12:44.960 --> 12:46.400
Initially money, not necessarily your boss,
12:46.400 --> 12:47.880
but like your peers saying, wow,
12:47.880 --> 12:50.400
this person really knows their stuff has a lot of value.
12:50.400 --> 12:53.280
So the reputation system came from that.
12:53.280 --> 12:56.320
So we were sort of Frankensteining a bunch of stuff together
12:56.320 --> 12:58.680
in Stack Overflow, like stuff we had seen working
12:58.680 --> 13:03.680
and we knew worked and that became Stack Overflow.
13:03.760 --> 13:06.520
And over time we realized it wasn't really discussion.
13:06.520 --> 13:08.760
It was very focused questions and answers.
13:08.760 --> 13:10.840
There wasn't a lot of room on the page for,
13:10.840 --> 13:12.560
let me talk about this tangential thing.
13:12.560 --> 13:14.800
It was more like, okay, is it answering the question?
13:14.800 --> 13:16.160
Is it clarifying the question
13:16.160 --> 13:18.720
or could it be an alternative answer to the same question?
13:18.720 --> 13:21.040
Cause there's usually more than one way to do it in program.
13:21.040 --> 13:22.680
There's like, say five to 10 ways.
13:22.680 --> 13:24.360
And one of the patterns we got into early on
13:24.360 --> 13:26.440
in Stack Overflow was there were questions
13:26.440 --> 13:28.440
where there would be like hundreds of answers.
13:28.440 --> 13:30.160
And we're like, well,
13:30.160 --> 13:31.840
how can there be a programming question
13:31.840 --> 13:34.760
with 100, 200, 500 answers?
13:34.760 --> 13:36.160
And we looked at those and we realized
13:36.160 --> 13:38.840
those were not really questions in the traditional sense.
13:38.840 --> 13:39.800
They were discussions.
13:39.800 --> 13:42.280
It was stuff that we allowed early on
13:42.280 --> 13:43.840
that we eventually decided wasn't allowed,
13:43.840 --> 13:46.760
such as what's your favorite programming food?
13:46.760 --> 13:49.280
What's the funniest programming cartoon you've seen?
13:49.280 --> 13:51.320
And we had to sort of backfill a bunch of rules
13:51.320 --> 13:53.080
about like, why isn't this allowed?
13:53.080 --> 13:55.160
Such as, is this a real problem you're facing?
13:55.160 --> 13:56.000
Like nobody goes to work
13:56.000 --> 13:57.200
and says, wow, I can't work
13:57.200 --> 13:59.080
cause I don't know what the funniest programming cartoon is.
13:59.080 --> 14:02.040
So sorry, can't compile this code now, right?
14:02.040 --> 14:04.600
It's not a real problem you're facing in your job.
14:04.600 --> 14:05.440
So that was Run Rule.
14:05.440 --> 14:07.320
And the second, like, what can you really learn from that?
14:07.320 --> 14:08.760
It's like what I call accidental learning
14:08.760 --> 14:10.200
or Reddit style learning
14:10.200 --> 14:11.640
where you're just like, oh, I'll just browse some things
14:11.640 --> 14:12.720
and oh, wow, you know,
14:12.720 --> 14:14.920
did you know tree frogs only live three years?
14:14.920 --> 14:16.000
I mean, I just made that up.
14:16.000 --> 14:16.920
I don't know if that's true,
14:16.920 --> 14:18.920
but I didn't really set out to learn that.
14:18.920 --> 14:20.280
I don't need to know that, right?
14:20.280 --> 14:21.720
It's accidental learning.
14:21.720 --> 14:23.440
It was more intentional learning
14:23.440 --> 14:24.600
where you're like, okay, I have a problem
14:24.600 --> 14:26.880
and I wanna learn about stuff around this problem having,
14:26.880 --> 14:27.720
right?
14:27.720 --> 14:29.160
And it could be theory, it could be compiler theory,
14:29.160 --> 14:29.960
it could be other stuff,
14:29.960 --> 14:31.800
but I'm having a compiler problem,
14:31.800 --> 14:34.280
hence I need to know the compiler theory,
14:34.280 --> 14:38.120
that aspect of it that gets me to my answer, right?
14:38.120 --> 14:39.360
So kind of a directed learning.
14:39.360 --> 14:40.920
So we had to backfill all these rules
14:40.920 --> 14:43.640
as we sort of figured out what the heck it was we were doing.
14:43.640 --> 14:45.360
And the system came very strict over time
14:45.360 --> 14:46.840
and a lot of people still complain about that.
14:46.840 --> 14:49.080
And I wrote my latest blog entry,
14:49.080 --> 14:51.000
what does Stack Overflow want to be?
14:51.000 --> 14:52.400
What does it want to be when it grows up?
14:52.400 --> 14:54.240
Celebrating the 10 year anniversary, yeah.
14:54.240 --> 14:56.320
Yeah, so 10 years and the system
14:56.320 --> 14:57.760
has trended towards strictness.
14:57.760 --> 14:58.960
There's a variety of reasons for this.
14:58.960 --> 15:01.400
One is people don't like to see other people
15:01.400 --> 15:04.040
get reputation for stuff as they view as frivolous,
15:04.040 --> 15:05.160
which I can actually understand
15:05.160 --> 15:07.920
because if you saw a programmer got like 500 upvotes
15:07.920 --> 15:10.440
for funniest programming cartoon
15:10.440 --> 15:11.840
or funniest comment they had seen in code.
15:11.840 --> 15:13.880
It's like, well, why do they have that reputation?
15:13.880 --> 15:15.120
Is it because they wrote the joke?
15:15.120 --> 15:15.960
Probably not.
15:15.960 --> 15:18.320
I mean, if they did, maybe, or the cartoon, right?
15:18.320 --> 15:19.520
They're getting a bunch of reputation
15:19.520 --> 15:20.800
based on someone else's work.
15:20.800 --> 15:22.720
It's not even like programming.
15:22.720 --> 15:23.640
It's just a joke, right?
15:23.640 --> 15:24.840
It's a related program.
15:24.840 --> 15:26.480
So you begin to resent that.
15:26.480 --> 15:27.560
You're like, well, that's not fair.
15:27.560 --> 15:28.400
And it isn't.
15:28.400 --> 15:29.240
At some level, they're correct.
15:29.240 --> 15:31.280
I empathize because it's not correct
15:31.280 --> 15:32.120
to get reputation for that.
15:32.120 --> 15:36.080
Versus here's a really gnarly, regular expression problem.
15:36.080 --> 15:40.520
And here's a really clever, insightful, detailed answer
15:40.520 --> 15:42.240
laying out, oh, here's why you're seeing the behavior
15:42.240 --> 15:43.080
that you're seeing.
15:43.080 --> 15:43.920
And here, let me teach you some things
15:43.920 --> 15:44.760
about how to avoid that in the future.
15:44.760 --> 15:46.560
That's great, that's gold, right?
15:46.560 --> 15:48.000
You want people to get reputation for that,
15:48.000 --> 15:51.040
not so much for, wow, look at this funny thing I saw, right?
15:51.040 --> 15:54.400
Great, so there's this very specific Q&A format,
15:54.400 --> 15:56.680
and then take me through the journey
15:56.680 --> 15:58.680
towards discourse in Facebook and Twitter.
15:58.680 --> 16:00.520
So you start at the beginning,
16:00.520 --> 16:03.040
that Stack Overflow evolved to have a purpose.
16:03.040 --> 16:06.800
So what is discourse, this passion you have
16:06.800 --> 16:09.840
for creating community for discussion?
16:09.840 --> 16:12.200
What is that, when was that born?
16:12.200 --> 16:13.680
Well, part of it is based on the realization
16:13.680 --> 16:15.040
that Stack Overflow is only good
16:15.040 --> 16:17.680
for very specific subjects where there's sort of,
16:17.680 --> 16:19.760
it's based on data, facts, and science,
16:19.760 --> 16:22.560
where answers can be kind of verified to be true.
16:22.560 --> 16:24.920
Another form of that is there's the book of knowledge,
16:24.920 --> 16:28.760
like the tome of knowledge that defines like whatever it is,
16:28.760 --> 16:30.920
you can refer to that book and it'll give you the answer.
16:30.920 --> 16:33.160
There has to be, it only works on subjects
16:33.160 --> 16:35.400
where there's like semi clear answers to things
16:35.400 --> 16:37.880
that can be verified in some form.
16:37.880 --> 16:39.840
Now again, there's always more than one way to do it.
16:39.840 --> 16:42.000
There's complete flexibility and system around that.
16:42.000 --> 16:45.360
But where it falls down is stuff like poker and Lego.
16:45.360 --> 16:48.800
Like we had, if you go to stackexchange.com,
16:48.800 --> 16:50.480
we have an engine that tries to launch
16:50.480 --> 16:52.360
different Q&A topics, right?
16:52.360 --> 16:57.040
And people can propose Q&A topics, sample questions,
16:57.040 --> 16:59.120
and if it gets enough support within the network,
16:59.120 --> 17:00.160
we launch that Q&A site.
17:00.160 --> 17:02.360
So some of the ones we launched were poker and Lego,
17:02.360 --> 17:03.440
and they did horribly, right?
17:03.440 --> 17:05.720
Because, I mean, they might still be there
17:05.720 --> 17:07.880
lingering on in some form, but it was an experiment.
17:07.880 --> 17:09.080
This is like a test, right?
17:09.080 --> 17:11.720
And some subjects work super well in the stack engine,
17:11.720 --> 17:12.840
and some don't.
17:12.840 --> 17:14.680
But the reason Lego and poker don't work
17:14.680 --> 17:16.080
is because they're so social, really.
17:16.080 --> 17:19.120
It's not about what's the rule here in poker.
17:19.120 --> 17:21.480
It's like, well, what kind of cigars
17:21.480 --> 17:23.520
do we like to smoke while playing poker?
17:23.520 --> 17:26.960
Or what's a cool set of cards to use when I'm playing poker?
17:26.960 --> 17:28.760
Or what's some strategies?
17:28.760 --> 17:30.120
Like say I have this hand come up,
17:30.120 --> 17:31.280
what's some strategies I could use?
17:31.280 --> 17:33.120
It's more of a discussion around what's happening.
17:33.120 --> 17:34.960
Like with Lego, same thing, like here's
17:34.960 --> 17:36.040
this cool Lego set I found.
17:36.040 --> 17:36.920
Look how awesome this is.
17:36.920 --> 17:38.600
And I'm like, yeah, that's freaking awesome, right?
17:38.600 --> 17:39.920
It's not a question, right?
17:39.920 --> 17:41.840
There's all these social components to the discussions
17:41.840 --> 17:43.080
that don't fit at all.
17:43.080 --> 17:45.400
Like we literally have to disallow those in Stack Overflow
17:45.400 --> 17:46.560
because it's not about being social.
17:46.560 --> 17:49.440
It's about problems that you're facing in your work
17:49.440 --> 17:51.400
that you need concrete answers for, right?
17:51.400 --> 17:52.840
Like you have a real demonstrated problem
17:52.840 --> 17:54.000
that's sort of blocking you and something.
17:54.000 --> 17:56.520
Nobody's blocked by what should I do
17:56.520 --> 17:58.040
when I have a straight flush, right?
17:58.040 --> 18:00.520
Like it's not a blocking problem in the world.
18:00.520 --> 18:02.400
It's just an opportunity to hang out and discuss.
18:02.400 --> 18:05.920
So discourse was a way to address that and say, look,
18:05.920 --> 18:10.800
you know, discussion form software was very, very bad.
18:10.800 --> 18:13.560
And when I came out of Stack Overflow in late 20,
18:13.560 --> 18:18.480
early 20, early 2013, early 2012,
18:18.480 --> 18:19.880
it was still very, very bad.
18:19.880 --> 18:22.120
I expected it improved in the four years
18:22.120 --> 18:24.600
since I last looked, but it had not improved at all.
18:24.600 --> 18:26.000
And I was like, well, that's kind of terrible
18:26.000 --> 18:29.040
because I love these communities of people
18:29.040 --> 18:30.920
talking about things that they love, you know,
18:30.920 --> 18:32.920
that there's just communities of interest, right?
18:32.920 --> 18:34.600
And there's no good software for them.
18:34.600 --> 18:37.440
Like startups would come to me and say, hey, Jeff,
18:37.440 --> 18:39.760
I wanna, you know, I have this startup.
18:39.760 --> 18:41.080
Here's my idea.
18:41.080 --> 18:42.960
And the first thing I would say to them is like, well,
18:42.960 --> 18:44.000
first, why are you asking me?
18:44.000 --> 18:46.440
Like I don't really know your field, right?
18:46.440 --> 18:48.320
Linus, sincerely, like, why aren't you asking
18:48.320 --> 18:49.760
like the community, like the people
18:49.760 --> 18:51.280
that are interested in this problem,
18:51.280 --> 18:52.480
the people that are using your product,
18:52.480 --> 18:53.840
why aren't you talking to them?
18:53.840 --> 18:55.200
And then they'd say, oh, great idea.
18:55.200 --> 18:56.120
Like how do I do that?
18:56.120 --> 18:58.120
And then that's when I started playing sad trombone
18:58.120 --> 19:00.080
because I realized all the software involving
19:00.080 --> 19:03.320
talking to your users, customers, audience, patrons,
19:03.320 --> 19:05.200
whatever it is, it was all really bad.
19:05.200 --> 19:06.680
You know, it was like stuff that I would be embarrassed
19:06.680 --> 19:08.240
to recommend to other people.
19:08.240 --> 19:09.960
And yet that's where I felt they could get
19:09.960 --> 19:12.800
the biggest and strongest, most effective input
19:12.800 --> 19:15.600
for what they should be doing with their product, right?
19:15.600 --> 19:17.520
It's from their users, from their community, right?
19:17.520 --> 19:18.880
That's what we did on Stack Overflow.
19:18.880 --> 19:21.360
So what we're talking about with forums,
19:21.360 --> 19:25.480
the, what is it, the dark matter of the internet,
19:25.480 --> 19:26.840
it's still, I don't know if it's still,
19:26.840 --> 19:30.920
but for the longest time, it has some of the most
19:30.920 --> 19:32.800
passionate and fascinating discussions.
19:32.800 --> 19:34.640
And what's the usual structure?
19:34.640 --> 19:37.000
There's usually what, it's linear.
19:37.000 --> 19:39.640
So it's sequential, so you're posting one after the other
19:39.640 --> 19:42.800
and there's Paging Nation, so it's every,
19:42.800 --> 19:45.040
there's a 10 posts and you go to the next page
19:45.040 --> 19:49.080
and that format still is used by, like I'm,
19:49.080 --> 19:51.480
we're doing a lot of research with Tesla of vehicles
19:51.480 --> 19:53.600
and there's a Tesla Motors Club forum,
19:53.600 --> 19:54.440
which is extremely,
19:54.440 --> 19:56.280
We really wanted to run that actually.
19:56.280 --> 19:57.720
They pinged us about it, I don't think we got it,
19:57.720 --> 19:59.240
but I really would have liked to gotten that one.
19:59.240 --> 20:02.640
But they've started before even 2012, I believe.
20:02.640 --> 20:04.200
I mean, they've been running for a long time.
20:04.200 --> 20:06.840
It's still an extremely rich source of information.
20:06.840 --> 20:09.760
So what's broken about that system
20:09.760 --> 20:12.160
and how are you trying to fix it?
20:12.160 --> 20:15.920
I think there's a lot of power in connecting people
20:15.920 --> 20:19.240
that love the same stuff around that specific topic,
20:19.240 --> 20:22.800
meaning Facebook's idea of connection is just any human
20:22.800 --> 20:24.560
that's related to another human, right?
20:24.560 --> 20:27.520
Like through friendship or any other reason.
20:27.520 --> 20:30.280
Facebook's idea of the world is sort of the status update,
20:30.280 --> 20:31.120
right?
20:31.120 --> 20:35.040
Like a friend of yours did something at a restaurant, right?
20:35.040 --> 20:37.280
Whereas discussion forums were traditionally
20:37.280 --> 20:38.320
around the interest graph.
20:38.320 --> 20:42.000
Like I love electric cars, specifically I love Tesla, right?
20:42.000 --> 20:44.320
Like I love the way they approach the problem.
20:44.320 --> 20:45.920
I love the style of the founder.
20:45.920 --> 20:48.200
I just love the design ethic.
20:48.200 --> 20:49.280
There's a lot to like about Tesla.
20:49.280 --> 20:50.280
I don't know if you saw the oatmeal,
20:50.280 --> 20:53.120
he did a whole love comic to Tesla.
20:53.120 --> 20:53.960
And it was actually kind of cool
20:53.960 --> 20:54.800
because I learned some stuff.
20:54.800 --> 20:57.000
He was talking about how great Tesla cars were specifically,
20:57.000 --> 20:58.360
like how they were built differently.
20:58.360 --> 20:59.880
And he went into a lot of great detail
20:59.880 --> 21:00.720
that was really interesting.
21:00.720 --> 21:02.640
And to me, that oatmeal post, if you read it,
21:02.640 --> 21:05.760
is the genesis of pretty much all interest communities.
21:05.760 --> 21:07.000
I just really love this stuff.
21:07.000 --> 21:08.400
So for me, for example, there's yoyos, right?
21:08.400 --> 21:09.720
Like I'm into the yoyo communities
21:09.720 --> 21:12.800
and these interest communities are just really fascinating
21:12.800 --> 21:15.240
to me and I feel more connected to the yoyo communities
21:15.240 --> 21:18.640
than I do to friends that I don't see that often, right?
21:18.640 --> 21:21.760
Like to me, the powerful thing is the interest graph
21:21.760 --> 21:25.040
and Facebook kind of dabbles in the interest graph.
21:25.040 --> 21:26.720
I mean, they have groups, you can sign up for groups
21:26.720 --> 21:29.480
and stuff, but it's really about the relationship graph.
21:29.480 --> 21:32.080
Like this is my coworker, this is my relative,
21:32.080 --> 21:35.280
this is my friend, but not so much about the interest.
21:35.280 --> 21:36.920
So I think that's the linchpin
21:36.920 --> 21:39.200
of which forums and communities are built on
21:39.200 --> 21:40.320
that I personally love.
21:40.320 --> 21:44.240
Like I said, leadership is about passion, right?
21:44.240 --> 21:45.520
And being passionate about stuff
21:45.520 --> 21:47.520
is a really valid way to look at the world.
21:47.520 --> 21:48.680
And I think it's a way,
21:49.800 --> 21:51.240
a lot of stuff in the world gets done.
21:51.240 --> 21:52.960
Like I once had someone describe me as,
21:52.960 --> 21:54.640
he's like, Jeff, you're a guy who,
21:54.640 --> 21:57.560
you just get super passionate about a few things at a time
21:57.560 --> 21:59.760
and you just go super deep in those things.
21:59.760 --> 22:00.720
And I was like, oh, that's kind of right.
22:00.720 --> 22:01.560
That's kind of what I do.
22:01.560 --> 22:03.400
I get into something and just be super into that
22:03.400 --> 22:04.560
for a couple of years or whatever
22:04.560 --> 22:05.960
and just learn all I can about it
22:05.960 --> 22:07.800
and go super deep in it.
22:07.800 --> 22:11.000
And that's how I enjoy experiencing the world, right?
22:11.000 --> 22:12.400
Like not being shallow on a bunch of things,
22:12.400 --> 22:15.360
but being really deep on a few things that I'm interested in.
22:15.360 --> 22:17.200
So forums kind of unlock that, right?
22:17.200 --> 22:19.680
And you don't want a world where everything belongs
22:19.680 --> 22:20.920
to Facebook, at least I don't.
22:20.920 --> 22:22.760
I want a world where communities can kind of own themselves,
22:22.760 --> 22:25.120
set their own norms, set their own rules,
22:25.120 --> 22:26.320
control the experience.
22:26.320 --> 22:28.720
Because community is also about ownership, right?
22:28.720 --> 22:31.520
Like if you're meeting at the Barnes & Noble
22:31.520 --> 22:33.000
every Thursday and Barnes & Noble says,
22:33.000 --> 22:34.600
get out of here, you guys don't buy enough books,
22:34.600 --> 22:36.640
well, you're kind of hoes, right?
22:36.640 --> 22:37.880
Barnes & Noble owns you, right?
22:37.880 --> 22:38.880
Like you can't.
22:38.880 --> 22:40.320
But if you have your own meeting space,
22:40.320 --> 22:41.440
you know, your own clubhouse,
22:41.440 --> 22:43.040
you can set your own rules,
22:43.040 --> 22:44.520
decide what you want to talk about there
22:44.520 --> 22:47.800
and just really generate a lot better information
22:47.800 --> 22:49.520
than you could like hanging out at Barnes & Noble
22:49.520 --> 22:51.880
every Thursday at 3 p.m., right?
22:51.880 --> 22:53.840
So that's kind of the vision of Discourse,
22:53.840 --> 22:57.160
is a place where it's fully open source.
22:57.160 --> 22:59.160
You can take the software, you can install it anywhere.
22:59.160 --> 23:00.880
And you know, you and a group of people
23:00.880 --> 23:02.840
can go deep on whatever it is that you're into.
23:02.840 --> 23:04.440
And this works for startups, right?
23:04.440 --> 23:05.920
Startups are a group of people
23:05.920 --> 23:08.400
who go super deep on a specific problem, right?
23:08.400 --> 23:09.600
And they want to talk to their communities
23:09.600 --> 23:11.040
like, well, install Discourse, right?
23:11.040 --> 23:12.040
That's what we do at Discourse.
23:12.040 --> 23:13.560
That's what I did at Stack Overflow.
23:13.560 --> 23:15.520
I spent a lot of time on MetaStack Overflow,
23:15.520 --> 23:17.920
which is our internal, well,
23:17.920 --> 23:20.200
public community feedback site.
23:20.200 --> 23:24.160
And just experiencing what the users were experiencing, right?
23:24.160 --> 23:26.040
Because they're the ones doing all the work in the system.
23:26.040 --> 23:27.600
And they had a lot of interesting feedback.
23:27.600 --> 23:30.200
And there's that 90, 10 rule of like 90% of the feedback
23:30.200 --> 23:32.360
you get is not really actionable for a variety of reasons.
23:32.360 --> 23:33.560
It might be bad feedback.
23:33.560 --> 23:34.600
It might be crazy feedback.
23:34.600 --> 23:36.400
It might be feedback you just can't act on right now.
23:36.400 --> 23:38.320
But there's 10% of it that's like gold.
23:38.320 --> 23:39.760
It's like literally gold and diamonds,
23:39.760 --> 23:42.440
where it's like feedback of really good improvements
23:42.440 --> 23:44.800
to your core product that are not super hard to get to
23:44.800 --> 23:45.800
and actually make a lot of sense.
23:45.800 --> 23:47.800
And my favorite is about 5% of those stuff
23:47.800 --> 23:48.720
I didn't even see coming.
23:48.720 --> 23:50.480
It's like, oh my God, I never even thought of that.
23:50.480 --> 23:52.560
But that's a brilliant idea, right?
23:52.560 --> 23:54.560
And I can point to so many features of Stack Overflow
23:54.560 --> 23:56.720
that we drive from MetaStack Overflow feedback
23:56.720 --> 24:00.280
and MetaDiscourse, right, same exact principle of discourse.
24:00.280 --> 24:02.280
You know, we're getting ideas from the community.
24:02.280 --> 24:03.480
I was like, oh my God, I never thought of that.
24:03.480 --> 24:04.680
But that's fantastic, right?
24:04.680 --> 24:06.920
Like I love that relationship with the community.
24:06.920 --> 24:10.200
From having built these communities, what have you,
24:10.200 --> 24:11.120
what have you learned about?
24:11.120 --> 24:12.880
What's the process of getting a critical mass
24:12.880 --> 24:14.160
of members in a community?
24:14.160 --> 24:17.040
Is it luck, skill, timing, persistence?
24:17.040 --> 24:20.400
What is, is it the tools, like discourse
24:20.400 --> 24:21.880
that empower that community?
24:21.880 --> 24:25.880
What's the key aspect of starting for one guy or gal
24:25.880 --> 24:27.760
and then building it to two, one and 10
24:27.760 --> 24:30.000
and a hundred and a thousand and so on?
24:30.000 --> 24:32.320
I think we're starting with an end of one.
24:32.320 --> 24:33.920
I mean, I think it's persistence
24:33.920 --> 24:37.040
and also you have to be interesting.
24:37.040 --> 24:38.520
Like somebody I really admire
24:38.520 --> 24:40.200
once said something that I always liked about blogging.
24:40.200 --> 24:41.880
He's like, here's how you blog.
24:41.880 --> 24:44.440
You have to have something interesting to say
24:44.440 --> 24:46.960
and have an interesting way of saying it, right?
24:46.960 --> 24:49.000
And then do that for like 10 years.
24:49.920 --> 24:52.360
So that's the genesis is like you have to have
24:52.360 --> 24:53.560
sort of something interesting to say
24:53.560 --> 24:55.280
that's not exactly what everybody else is saying.
24:55.280 --> 24:56.440
And an interesting way of saying it,
24:56.440 --> 24:57.920
which is another way of saying kind of entertaining
24:57.920 --> 24:59.280
way of saying it.
24:59.280 --> 25:02.200
And then as far as growing it, it's like ritual, you know,
25:02.200 --> 25:04.280
like you have to like say you're starting a blog,
25:04.280 --> 25:06.720
you have to say, look, I'm gonna blog every week,
25:06.720 --> 25:09.320
three times a week and you have to stick to that schedule,
25:09.320 --> 25:10.160
right?
25:10.160 --> 25:12.880
Because until you do that for like several years,
25:12.880 --> 25:14.400
you're never gonna get anywhere.
25:14.400 --> 25:16.720
Like it just takes years to get to where you need to get to.
25:16.720 --> 25:18.720
And part of that is having the discipline
25:18.720 --> 25:19.760
to stick with the schedule.
25:19.760 --> 25:21.720
And it helps again, if it's something you're passionate about,
25:21.720 --> 25:22.760
this won't feel like work.
25:22.760 --> 25:23.600
Like, I love this.
25:23.600 --> 25:25.640
I can talk about this all day, every day, right?
25:26.840 --> 25:28.000
You just have to do it in a way that's interesting
25:28.000 --> 25:28.840
to other people.
25:28.840 --> 25:30.640
And then as you're growing the community,
25:30.640 --> 25:32.480
that pattern of participation within the community
25:32.480 --> 25:34.320
of like generating these artifacts
25:34.320 --> 25:35.720
and inviting other people to help you
25:35.720 --> 25:36.960
like collaborate on these artifacts.
25:36.960 --> 25:38.280
Like even in the case of blogging,
25:38.280 --> 25:40.880
like I felt in the early days of my blog,
25:40.880 --> 25:41.960
which I started in 2004,
25:41.960 --> 25:43.760
which is really the genesis of Stack Overflow.
25:43.760 --> 25:46.280
If you look at all my blog, it leads up to Stack Overflow,
25:46.280 --> 25:48.440
which was, I have all this energy in my blog,
25:48.440 --> 25:51.320
but I don't like 40,000 people were subscribing to me.
25:51.320 --> 25:52.440
And I was like, I wanna do something.
25:52.440 --> 25:54.160
And then I met Joel and said, hey, Joel,
25:54.160 --> 25:56.240
I wanna do something, take this ball of energy
25:56.240 --> 25:57.360
from my blog and do something.
25:57.360 --> 25:58.800
And all the people who read my blog saw that,
25:58.800 --> 26:00.560
it's like, oh, cool, you're involving us.
26:00.560 --> 26:03.360
You're saying, look, you're part of this community,
26:03.360 --> 26:04.520
let's build this thing together.
26:04.520 --> 26:05.520
Like they pick the name.
26:05.520 --> 26:08.040
Like we voted on the name for Stack Overflow on my blog.
26:08.040 --> 26:09.960
Like we came up and naming is super hard.
26:09.960 --> 26:12.240
First of all, the hardest problem in computer science
26:12.240 --> 26:14.480
is coming with a good name for stuff, right?
26:14.480 --> 26:15.880
But you can go back to my blog.
26:15.880 --> 26:17.600
There's the poll where we voted
26:17.600 --> 26:19.240
and Stack Overflow became the name of the site.
26:19.240 --> 26:21.160
And all the early beta users of Stack Overflow
26:21.160 --> 26:24.560
were audience of my blog plus Joel's blog, right?
26:24.560 --> 26:26.480
So we started from like, if you look at the Genesis,
26:26.480 --> 26:28.120
okay, I was just a programmer who said,
26:28.120 --> 26:30.400
hey, I love programming, but I have no outlet
26:30.400 --> 26:31.240
to talk about it.
26:31.240 --> 26:32.080
So I'm just gonna blog about it
26:32.080 --> 26:33.320
because I don't have enough people to work
26:33.320 --> 26:34.320
to talk to about it.
26:34.320 --> 26:36.480
Cause at the time I worked place where,
26:36.480 --> 26:38.880
you know, programming wasn't the core output of the company.
26:38.880 --> 26:40.800
It was a pharmaceutical company.
26:40.800 --> 26:43.520
And I just love this stuff, you know, to an absurd degree.
26:43.520 --> 26:44.920
So I was like, I'll just blog about it
26:44.920 --> 26:45.920
and then I'll find an audience
26:45.920 --> 26:47.680
and eventually found an audience,
26:47.680 --> 26:50.440
eventually found Joel and eventually built Stack Overflow
26:50.440 --> 26:52.840
from that one core of activity, right?
26:52.840 --> 26:55.400
But it was that repetition of feeding back in,
26:55.400 --> 26:58.040
feedback from my blog comments, feedback from Joel,
26:58.040 --> 27:01.440
feedback from the early Stack Overflow community.
27:01.440 --> 27:02.840
When people see that you're doing that,
27:02.840 --> 27:04.120
they will follow along with you, right?
27:04.120 --> 27:05.480
They say, cool, you're here in good faith.
27:05.480 --> 27:07.240
You're actually, you know, not listening to everything
27:07.240 --> 27:09.280
because that's impossible, that's impossible,
27:09.280 --> 27:13.000
but you're actually, you know, waiting our feedback
27:13.000 --> 27:14.720
in what you're doing because, and why wouldn't I?
27:14.720 --> 27:16.680
Because who does all the work on Stack Overflow?
27:16.680 --> 27:17.920
Me, Joel?
27:17.920 --> 27:19.600
No, it's the other programmers
27:19.600 --> 27:20.480
that are doing all the work.
27:20.480 --> 27:22.480
So you gotta have some respect for that.
27:22.480 --> 27:25.080
And then, you know, discipline around, look,
27:25.080 --> 27:26.400
you know, we're trying to do a very specific thing
27:26.400 --> 27:27.240
here on Stack Overflow.
27:27.240 --> 27:29.080
We're not trying to solve all the world's problems.
27:29.080 --> 27:30.920
We're trying to solve this very specific Q&A problem
27:30.920 --> 27:32.320
in a very specific way.
27:32.320 --> 27:33.560
Not because we're jerks about it,
27:33.560 --> 27:35.960
but because these strict set of rules
27:35.960 --> 27:39.360
help us get really good results, right?
27:39.360 --> 27:41.440
And programmers, that's an easy sell for the most part
27:41.440 --> 27:42.520
because programmers are used to dealing
27:42.520 --> 27:45.080
with ridiculous systems of rules like constantly.
27:45.080 --> 27:46.760
That's basically their job.
27:46.760 --> 27:49.480
So they're very, oh yeah, super strict system of rules
27:49.480 --> 27:50.400
that lets me get what I want.
27:50.400 --> 27:51.560
That's programming, right?
27:51.560 --> 27:53.640
That's what Stack Overflow is, so.
27:53.640 --> 27:54.880
So you're making it sound easy,
27:54.880 --> 27:58.000
but in 2004, let's go back there.
27:58.000 --> 28:01.520
In 2004, you started the blog, Coding Horror.
28:01.520 --> 28:03.520
Was it called that at the very beginning?
28:03.520 --> 28:04.360
It was.
28:04.360 --> 28:05.200
One of the smart things I did,
28:05.200 --> 28:06.680
it's from a book by Steve McConnell, Code Complete,
28:06.680 --> 28:08.200
which is one of my favorite programming books,
28:08.200 --> 28:09.760
still probably my number one programming book
28:09.760 --> 28:10.760
for anyone to read.
28:12.080 --> 28:13.800
One of the smart things I did back then,
28:13.800 --> 28:16.040
I don't always do smart things when I start stuff.
28:16.040 --> 28:18.360
I contacted Steve and said, hey, I really like this.
28:18.360 --> 28:19.600
It was a sidebar illustration,
28:19.600 --> 28:21.800
indicating danger in code, right?
28:21.800 --> 28:23.880
Coding Horror was like, watch out.
28:25.280 --> 28:27.400
And I love that illustration because it spoke to me.
28:27.400 --> 28:28.840
Because I saw that illustration go, oh my God,
28:28.840 --> 28:31.040
that's me, like I'm always my own worst enemy.
28:31.040 --> 28:33.320
Like that's the key insight in programming is,
28:33.320 --> 28:34.280
every time you write something,
28:34.280 --> 28:36.560
think how am I gonna screw myself?
28:36.560 --> 28:39.240
Because you will, constantly, right?
28:39.240 --> 28:41.240
So that icon was like, oh yeah,
28:41.240 --> 28:43.160
I need to constantly hold that mirror up and look
28:43.160 --> 28:45.400
and say, look, you're very fallible.
28:45.400 --> 28:46.360
You're gonna screw this up.
28:46.360 --> 28:48.000
Like how can you build this in such a way
28:48.000 --> 28:50.320
that you're not gonna screw it up later?
28:50.320 --> 28:52.960
Like how can you get that discipline around
28:52.960 --> 28:54.280
making sure at every step,
28:54.280 --> 28:56.160
I'm thinking through all the things that I could do wrong
28:56.160 --> 28:57.200
or that other people could do wrong.
28:57.200 --> 28:59.520
Cause that is actually how you get to be a better programmer
28:59.520 --> 29:00.560
a lot of times, right?
29:00.560 --> 29:03.360
So that sidebar illustration, I loved it so much.
29:03.360 --> 29:05.440
And I wrote Steve before I started my blog,
29:05.440 --> 29:06.760
and said, hey, can I have permission to use this?
29:06.760 --> 29:08.400
Cause I just really like this illustration.
29:08.400 --> 29:11.240
And Steve was kind enough to let me permission to do that
29:11.240 --> 29:12.640
and just continues to give me permission.
29:12.640 --> 29:13.480
So yeah.
29:13.480 --> 29:15.040
Really, that's awesome.
29:15.040 --> 29:18.840
But in 2004, you started this blog.
29:18.840 --> 29:22.560
You know, you look at Stephen King's book on writing
29:22.560 --> 29:26.400
or Stephen Pressfield's War of Art book.
29:26.400 --> 29:30.000
I mean, it seems like writers suffer.
29:30.000 --> 29:32.480
I mean, it's a hard process of writing, right?
29:32.480 --> 29:34.160
There's a lot of, there's gonna be suffering.
29:34.160 --> 29:35.160
I mean, I won't kid you like,
29:35.160 --> 29:36.360
well, the work is suffering, right?
29:36.360 --> 29:38.680
Like doing the work, like even when you're every week,
29:38.680 --> 29:40.240
you're like, okay, that blog post wasn't very good
29:40.240 --> 29:41.440
or you know, people didn't like it
29:41.440 --> 29:44.040
or people said disparaging things about it.
29:44.040 --> 29:47.400
You have to like have the attitude is like, you know,
29:47.400 --> 29:49.440
no matter what happens, I want to do this for me, right?
29:49.440 --> 29:51.200
It's not about you, it's about me.
29:51.200 --> 29:53.280
I mean, in the end, it is about everyone
29:53.280 --> 29:55.800
because this is how good work gets out into the world.
29:55.800 --> 29:59.760
But you have to be pretty strict about saying like,
29:59.760 --> 30:01.600
you know, I'm selfish in the sense
30:01.600 --> 30:03.240
that I have to do this for me.
30:03.240 --> 30:04.320
You know, you mentioned Stephen King,
30:04.320 --> 30:05.440
like his book on writing,
30:05.440 --> 30:06.800
but like one of the things I do, for example,
30:06.800 --> 30:08.640
when writing is like, I read it out loud,
30:08.640 --> 30:10.600
one of the best pieces of advice for writing anything
30:10.600 --> 30:13.840
is read it out loud, like multiple times.
30:13.840 --> 30:15.800
And make it sound like you're talking
30:15.800 --> 30:17.840
because that is the goal of good writing.
30:17.840 --> 30:19.520
It should sound like you said it with,
30:19.520 --> 30:20.840
with slightly better phrasing
30:20.840 --> 30:22.080
because you have more time to think about what you're saying,
30:22.080 --> 30:24.360
but like, it should sound natural when you say it.
30:24.360 --> 30:27.120
And I think that's probably the single best writing advice
30:27.120 --> 30:27.960
I can give anyone.
30:27.960 --> 30:29.760
It's just, just read it over and over out loud,
30:29.760 --> 30:32.600
make sure it sounds like something you would normally say
30:32.600 --> 30:33.680
and it sounds good.
30:33.680 --> 30:35.600
And what's your process of writing?
30:35.600 --> 30:39.360
So there's usually a pretty good idea behind the blog post.
30:39.360 --> 30:40.200
So ideas, right.
30:40.200 --> 30:44.480
So I think you gotta have the concept that
30:44.480 --> 30:46.160
there's so many interesting things in the world.
30:46.160 --> 30:48.760
Like, I mean, my God, the world is amazing, right?
30:48.760 --> 30:51.920
Like you could never write about everything that's going on
30:51.920 --> 30:52.760
because it's so incredible.
30:52.760 --> 30:54.160
But if you can't come up with like,
30:54.160 --> 30:57.080
let's say one interesting thing per day to talk about,
30:57.080 --> 30:58.320
then you're not trying hard enough
30:58.320 --> 31:00.600
because the world is full of just super interesting stuff.
31:00.600 --> 31:03.080
And one great way to like mine stuff
31:03.080 --> 31:04.240
is go back to old books
31:04.240 --> 31:07.160
because they bring up old stuff that's still super relevant.
31:07.160 --> 31:08.120
And I did that a lot
31:08.120 --> 31:09.840
because I was like reading classic programming books
31:09.840 --> 31:11.240
and a lot of the early blog posts were like,
31:11.240 --> 31:12.240
oh, I was reading this programming book
31:12.240 --> 31:13.720
and they brought this really cool concept
31:13.720 --> 31:14.960
and I want to talk about it some more.
31:14.960 --> 31:16.960
And you get the, I mean, you're not claiming credit
31:16.960 --> 31:18.080
for the idea, but it gives you something interesting
31:18.080 --> 31:19.840
to talk about that's kind of evergreen, right?
31:19.840 --> 31:21.920
Like you don't have to go, what should I talk about?
31:21.920 --> 31:24.560
So we'll just go dig up some old classic programming books
31:24.560 --> 31:26.960
and find something that, oh, wow, that's interesting.
31:26.960 --> 31:28.120
Or how does that apply today?
31:28.120 --> 31:30.920
Or what about X and Y or compare these two concepts?
31:30.920 --> 31:32.720
So pull a couple of sentences from that book
31:32.720 --> 31:35.960
and then sort of play off of it, almost agree or disagree.
31:35.960 --> 31:40.960
That so in 2007, you wrote that you were offered
31:41.880 --> 31:44.480
a significant amount of money to sell the blog.
31:44.480 --> 31:46.000
You chose not to.
31:46.000 --> 31:49.000
What were all the elements you were thinking about?
31:49.000 --> 31:50.560
Because I'd like to take you back.
31:50.560 --> 31:52.560
It seems like there's a lot of nonlinear decisions
31:52.560 --> 31:54.600
you made through life.
31:54.600 --> 31:56.080
So what was that decision like?
31:56.080 --> 31:57.880
Right, so one of the things I love
31:57.880 --> 31:59.400
is the Choose Your Own Adventure books,
31:59.400 --> 32:00.320
which I loved as a kid.
32:00.320 --> 32:01.760
And I feel like they're early programmer books
32:01.760 --> 32:04.080
because they're all about if then statements, right?
32:04.080 --> 32:07.280
If this then this, and they're also very, very unforgiving.
32:07.280 --> 32:08.960
Like there's all these sites that map
32:08.960 --> 32:11.080
the classic Choose Your Own Adventure books
32:11.080 --> 32:13.960
and how many outcomes are bad, a lot of bad outcomes.
32:13.960 --> 32:16.120
So part of the game is like, oh, I got an outcome.
32:16.120 --> 32:17.840
Go back one step, go back one further step.
32:17.840 --> 32:19.640
It's like, how did I get here, right?
32:19.640 --> 32:21.920
It's a sequence of decisions.
32:21.920 --> 32:23.200
And this is true of life, right?
32:23.200 --> 32:25.600
Like every decision is a sequence, right?
32:25.600 --> 32:28.360
Individually, any individual decision
32:28.360 --> 32:29.280
is not actually right or wrong,
32:29.280 --> 32:31.080
but they lead you down a path, right?
32:31.080 --> 32:32.800
So I do think there's some truth there.
32:32.800 --> 32:34.800
So this particular decision,
32:34.800 --> 32:36.360
the blogging I'm fairly popular.
32:36.360 --> 32:39.120
There's a lot of RSS readers that I discovered.
32:39.120 --> 32:41.040
And this guy contacted me out of the blue
32:41.040 --> 32:42.480
from this like bug tracking company.
32:42.480 --> 32:44.480
He's like, oh, I really want to buy your blog for like,
32:44.480 --> 32:47.040
I think it was around, it was $100,000,
32:47.040 --> 32:49.280
might've been like 80,000, but it was a lot, right?
32:49.280 --> 32:51.160
Like, and that's, you know, at the time,
32:51.160 --> 32:54.480
like I would have a year's worth of salary all at once.
32:54.480 --> 32:57.760
So I really think about like, well, you know,
32:57.760 --> 32:58.920
and I remember talking to people at the time,
32:58.920 --> 32:59.880
I was like, wow, that's a lot of money.
32:59.880 --> 33:02.040
But then the other thing, like I really like my blog, right?
33:02.040 --> 33:03.760
Like, do I want to sell my blog?
33:03.760 --> 33:05.320
Because it wouldn't really belong to me anymore
33:05.320 --> 33:06.160
at that point.
33:06.160 --> 33:09.400
And one of the guidelines that I like to,
33:09.400 --> 33:11.080
I don't like to give advice to people a lot,
33:11.080 --> 33:12.560
but one of the pieces of advice I do give,
33:12.560 --> 33:13.760
because I do think it's really true
33:13.760 --> 33:15.600
and it's generally helpful,
33:15.600 --> 33:17.640
is whenever you're looking at a set of decisions,
33:17.640 --> 33:19.920
like, oh gosh, should I do A, B or C?
33:19.920 --> 33:22.560
You got to pick the thing that's a little scarier
33:22.560 --> 33:24.560
in that list, because not, you know,
33:24.560 --> 33:25.720
not like jump off a cliff scary,
33:25.720 --> 33:27.040
but the thing that makes you nervous.
33:27.040 --> 33:29.000
Because if you pick the safe choice,
33:29.000 --> 33:30.360
it's usually you're not really pushing.
33:30.360 --> 33:31.400
You're not pushing yourself.
33:31.400 --> 33:33.960
You're not choosing the thing that's going to help you grow.
33:33.960 --> 33:36.160
So for me, the scarier choice was to say no.
33:36.160 --> 33:38.360
I was like, well, no, let's just see where this is going, right?
33:38.360 --> 33:39.520
Because then I own it.
33:39.520 --> 33:40.520
I mean, it belongs to me.
33:40.520 --> 33:41.760
It's my thing.
33:41.760 --> 33:43.160
And I can just take it
33:43.160 --> 33:45.360
and tell some other logical conclusion, right?
33:45.360 --> 33:46.840
Because imagine how different the world would have been
33:46.840 --> 33:49.080
had I said yes and sold the blog.
33:49.080 --> 33:51.560
It's like there probably wouldn't be Stack Overflow.
33:51.560 --> 33:53.640
You know, a lot of other stuff would have changed.
33:53.640 --> 33:55.360
So for that particular decision,
33:55.360 --> 33:56.320
I think it was that same rule,
33:56.320 --> 33:57.960
like what scares me a little bit more?
33:57.960 --> 33:59.240
Do the thing that scares you.
33:59.240 --> 34:00.080
Yeah.
34:00.080 --> 34:01.760
Which startups?
34:01.760 --> 34:03.280
I think there's a specific,
34:03.280 --> 34:05.000
some more general questions
34:05.000 --> 34:07.560
that a lot of people would be interested in.
34:07.560 --> 34:10.320
You've started Stack Overflow.
34:10.320 --> 34:12.080
You've started Discourse.
34:12.080 --> 34:15.320
So what's the, you know, one, two, three guys,
34:15.320 --> 34:17.200
whatever it is in the beginning.
34:17.200 --> 34:19.360
What was that process like?
34:19.360 --> 34:20.720
Do you start talking about it?
34:20.720 --> 34:21.800
Do you start programming?
34:21.800 --> 34:24.520
Do you start, like where is the birth
34:24.520 --> 34:25.840
and the catalyst that actually exists?
34:25.840 --> 34:27.000
Well, I can talk about it in the context
34:27.000 --> 34:28.080
of both Stack Overflow and Discourse.
34:28.080 --> 34:31.040
So I think the key thing initially is there is a problem.
34:31.040 --> 34:32.400
Something, there's some state of the world
34:32.400 --> 34:33.920
that's unsatisfactory to the point
34:33.920 --> 34:35.960
that like you're upset about it, right?
34:35.960 --> 34:37.800
Like in that case, it was experts exchange.
34:37.800 --> 34:38.920
I mean, Joel's original idea,
34:38.920 --> 34:40.560
cause I approached Joel's like, look, Joel,
34:40.560 --> 34:41.720
I have all this energy behind my blog.
34:41.720 --> 34:42.560
I want to do something.
34:42.560 --> 34:43.880
I want to build something, but I don't know what it is.
34:43.880 --> 34:45.680
Cause I'm not, I'm honestly not a good idea person.
34:45.680 --> 34:46.520
I'm really not.
34:46.520 --> 34:47.720
I'm like the execution guy.
34:47.720 --> 34:48.640
I'm really good at execution,
34:48.640 --> 34:51.000
but I'm not good at like blue skying ideas.
34:51.000 --> 34:52.040
Not my forte.
34:52.040 --> 34:54.040
Which is another reason why I like the community feedback.
34:54.040 --> 34:56.320
Cause they blue sky all day long for you, right?
34:56.320 --> 34:58.400
So when I can just go in and cherry pick a blue sky idea
34:58.400 --> 35:00.800
from community, even if I have to spend three hours reading
35:00.800 --> 35:02.720
to get one good idea, it's worth it, man.
35:02.720 --> 35:04.680
But anyway, so the idea from Joel was,
35:04.680 --> 35:07.400
hey, experts exchange, it's got great data,
35:07.400 --> 35:08.880
but the experience is hideous, right?
35:08.880 --> 35:09.920
It's, it's trying to trick you.
35:09.920 --> 35:12.040
It feels like you use car sales, but it's just bad.
35:12.040 --> 35:13.080
So I was like, oh, that's awesome.
35:13.080 --> 35:14.280
It feeds into community.
35:14.280 --> 35:15.120
It feeds into like, you know,
35:15.120 --> 35:16.680
we can make a creative commons.
35:16.680 --> 35:18.520
So I think the core is to have a really good idea
35:18.520 --> 35:20.280
that you feel very strongly about in the beginning,
35:20.280 --> 35:22.680
that like there's a wrong in the world that we will,
35:22.680 --> 35:24.280
an injustice that we will write
35:24.280 --> 35:26.120
through the process of building this thing.
35:26.120 --> 35:27.520
For discourse, it was like, look,
35:27.520 --> 35:30.200
there's no good software for communities
35:30.200 --> 35:32.960
to just hang out and like do stuff, right?
35:32.960 --> 35:35.480
Like whether it's problem solving, startup, whatever.
35:35.480 --> 35:37.640
Forms are such a great building block of online community
35:37.640 --> 35:38.600
and they're hideous.
35:38.600 --> 35:39.760
They were so bad, right?
35:39.760 --> 35:40.760
It was embarrassing.
35:40.760 --> 35:43.000
Like I literally was embarrassed to be associated
35:43.000 --> 35:43.960
with this software, right?
35:43.960 --> 35:45.720
I was like, we have to have software that you can be proud of.
35:45.720 --> 35:47.120
It's like, this is competitive with Reddit.
35:47.120 --> 35:48.360
This is competitive with Twitter.
35:48.360 --> 35:50.360
This is competitive with Facebook, right?
35:50.360 --> 35:53.520
I would be proud to have the software on my site.
35:53.520 --> 35:54.920
So that was the genesis of discourse,
35:54.920 --> 35:59.920
was feeling very strongly about their needs to be
35:59.920 --> 36:01.760
a good solution for communities.
36:01.760 --> 36:03.280
So that's step one, genesis funny idea
36:03.280 --> 36:04.600
you feel super strongly about, right?
36:04.600 --> 36:06.680
And then people galvanize around the idea.
36:06.680 --> 36:08.520
Like Joel was already super excited about the idea.
36:08.520 --> 36:10.000
I was excited about the idea.
36:10.000 --> 36:13.280
So with the forum software, I was posting on Twitter,
36:13.280 --> 36:15.040
I had researched, as part of my research,
36:15.040 --> 36:16.640
I start researching the problem, right?
36:16.640 --> 36:19.520
And I found a game called forum wars,
36:19.520 --> 36:21.480
which was a parody of forum.
36:21.480 --> 36:23.640
It's still very, very funny of like forum behavior
36:23.640 --> 36:26.120
a circle, like I would say 2003.
36:26.120 --> 36:27.720
It's age some, right?
36:27.720 --> 36:29.200
Like the behavior is a little different
36:29.200 --> 36:30.880
in the air of Twitter, but it was awesome.
36:30.880 --> 36:32.760
It was very funny and it was like a game,
36:32.760 --> 36:34.840
it was like an RPG and it had a forum attached to it.
36:34.840 --> 36:37.520
So it was like a game about forums with a forum attached.
36:37.520 --> 36:38.520
It was like, this is awesome, right?
36:38.520 --> 36:39.520
This is so cool.
36:39.520 --> 36:41.920
And the founder of that company or that project,
36:41.920 --> 36:43.040
it wasn't really a company,
36:43.040 --> 36:45.680
contacted me, this guy Robin Ward from Toronto.
36:45.680 --> 36:46.880
It's like, hey, you know, I saw you've been talking
36:46.880 --> 36:49.320
about forums and like, I really love that problem space.
36:49.320 --> 36:51.600
It's like, I'd still love to build really good forum software
36:51.600 --> 36:53.360
because I don't think anything out there is any good.
36:53.360 --> 36:54.200
And I was like, awesome.
36:54.200 --> 36:55.880
At that point, I was like, we're starting a company
36:55.880 --> 36:58.560
because like, I couldn't have wished for a better person
36:58.560 --> 37:00.080
to walk through the door and say,
37:00.080 --> 37:01.480
I'm excited about this too.
37:01.480 --> 37:02.360
Same thing with Joel, right?
37:02.360 --> 37:03.960
I mean, Joel is a legend in the industry, right?
37:03.960 --> 37:04.800
So when he walked through and said,
37:04.800 --> 37:05.640
I'm excited about this problem,
37:05.640 --> 37:08.160
I was like, me too, man, we can do this, right?
37:08.160 --> 37:10.600
So that to me is the most important step.
37:10.600 --> 37:12.040
It's like having an idea you're super excited about
37:12.040 --> 37:14.600
and another person, a co founder, right?
37:14.600 --> 37:16.200
Cause again, you get that dual leadership, right?
37:16.200 --> 37:18.080
Of like, am I making a bad decision?
37:19.080 --> 37:21.480
Sometimes it's nice to have checks of like,
37:21.480 --> 37:22.440
is this a good idea?
37:22.440 --> 37:23.680
I don't know, right?
37:23.680 --> 37:25.600
So those are the crucial seeds,
37:25.600 --> 37:27.520
but then starting to build stuff,
37:27.520 --> 37:28.360
whether it's you programming or somebody else.
37:28.360 --> 37:29.360
There is prototyping.
37:29.360 --> 37:30.600
So there's tons of research.
37:30.600 --> 37:32.920
There's tons of research, like what's out there that failed?
37:32.920 --> 37:34.360
Cause a lot of people look at it with successes.
37:34.360 --> 37:36.040
Oh, look at how successful X is.
37:36.040 --> 37:37.360
Everybody looks at the successes.
37:37.360 --> 37:38.360
Those are boring.
37:38.360 --> 37:39.560
Show me the failures.
37:39.560 --> 37:40.880
Cause that is what's interesting.
37:40.880 --> 37:42.080
That's where people were experimenting.
37:42.080 --> 37:43.520
That's where people were pushing,
37:43.520 --> 37:46.640
but they failed, but they probably failed for reasons
37:46.640 --> 37:50.080
that weren't directly about the quality of their idea, right?
37:50.080 --> 37:51.440
So look at all the failures.
37:51.440 --> 37:52.720
Don't just look what everybody looks at,
37:52.720 --> 37:53.560
which is like, oh gosh,
37:53.560 --> 37:54.680
look at all these successful people.
37:54.680 --> 37:55.960
Look at the failures.
37:55.960 --> 37:57.280
Look at the things that didn't work.
37:57.280 --> 37:59.080
Research the entire field.
37:59.080 --> 38:01.480
And so that's the research that I was doing
38:01.480 --> 38:02.760
that led me to Robin, right?
38:02.760 --> 38:03.600
Was that.
38:03.600 --> 38:04.800
And then when we, for example,
38:04.800 --> 38:07.560
when we did Stack Overflow,
38:07.560 --> 38:09.000
we're like, okay, well I really like elements
38:09.000 --> 38:10.040
of voting and digging Reddit.
38:10.040 --> 38:12.800
I like the Wikipedia, everything's up to date.
38:12.800 --> 38:14.320
Nothing is like an old tombstone
38:14.320 --> 38:16.980
that like has horrible out of date information.
38:16.980 --> 38:17.820
We know that works.
38:17.820 --> 38:19.400
Wikipedia is an amazing resource.
38:19.400 --> 38:22.400
Blogging the idea of ownership is so powerful, right?
38:22.400 --> 38:24.000
Like, oh, I Joe wrote this
38:24.000 --> 38:25.520
and look how good Joe's answer is, right?
38:25.520 --> 38:27.760
Like all these concepts were rolling together,
38:27.760 --> 38:29.080
researching all the things that were out there
38:29.080 --> 38:30.640
that were working and why they were working
38:30.640 --> 38:32.240
and trying to like fold them into that.
38:32.240 --> 38:34.040
Again, that Frankenstein's monster
38:34.040 --> 38:35.280
of what Stack Overflow is.
38:35.280 --> 38:37.080
And by the way, that wasn't a free decision
38:37.080 --> 38:38.760
because there's still a ton of tension
38:38.760 --> 38:40.040
in the Stack Overflow system.
38:40.040 --> 38:42.600
There's reasons people complain about Stack Overflow
38:42.600 --> 38:43.680
because it's so strict, right?
38:43.680 --> 38:44.640
Why is it so strict?
38:44.640 --> 38:46.480
Why are you guys always closing my questions?
38:46.480 --> 38:47.880
It's because there's so much tension
38:47.880 --> 38:49.120
that we built into the system
38:49.120 --> 38:51.240
around like trying to get good, good results
38:51.240 --> 38:56.000
out of the system and, you know, it's not a free.
38:56.000 --> 38:57.480
That stuff doesn't come for free, right?
38:57.480 --> 38:59.800
It's not like we all have perfect answers
38:59.800 --> 39:02.280
and nobody will have to get their feelings heard
39:02.280 --> 39:04.160
or nobody will have to get downvoted.
39:04.160 --> 39:05.800
Like it doesn't work that way, right?
39:05.800 --> 39:07.960
Like, so this is an interesting point
39:07.960 --> 39:09.400
and a small tangent.
39:09.400 --> 39:12.000
Yeah, you're right about anxiety.
39:12.000 --> 39:14.040
So I've posted a lot of questions
39:14.040 --> 39:16.040
and written answers on Stack Overflow.
39:16.040 --> 39:19.800
On the question side, usually go to something very specific
39:19.800 --> 39:21.200
to something I'm working on.
39:21.200 --> 39:22.280
This is something you talk about
39:22.280 --> 39:25.000
that really the goal of Stack Overflow
39:25.000 --> 39:27.840
isn't about, is to write a question not,
39:27.840 --> 39:31.800
that's not about you, it's about the question
39:31.800 --> 39:35.800
that will help the community in the future.
39:35.800 --> 39:37.520
Right, that's a tough sell, right?
39:37.520 --> 39:39.160
Because people are like, well, you know,
39:39.160 --> 39:40.280
I don't really care about the community.
39:40.280 --> 39:41.520
What I care about is my problem.
39:41.520 --> 39:42.400
My problem.
39:42.400 --> 39:43.480
And that's fair, right?
39:43.480 --> 39:44.960
It's sort of that, again, that tension,
39:44.960 --> 39:45.800
that balancing act.
39:45.800 --> 39:46.760
I want to help you,
39:46.760 --> 39:48.480
but we also want to help everybody that comes behind you.
39:48.480 --> 39:49.960
Right, the long line of people are gonna come up and say,
39:49.960 --> 39:52.320
oh, I kind of have that problem too, right?
39:52.320 --> 39:53.680
And if nobody's ever gonna come up
39:53.680 --> 39:54.880
and say I have this problem too,
39:54.880 --> 39:57.040
then that question shouldn't exist on Stack Overflow
39:57.040 --> 39:58.920
because the question is too specific.
39:58.920 --> 40:00.120
And even that's tension, right?
40:00.120 --> 40:00.960
How do you judge that?
40:00.960 --> 40:02.640
How do you know that nobody's ever gonna have
40:02.640 --> 40:04.720
this particular question again?
40:04.720 --> 40:06.840
So there's a lot of tension in the system.
40:06.840 --> 40:09.640
Do you think that anxiety of asking the question,
40:09.640 --> 40:12.680
the anxiety of answering that tension
40:12.680 --> 40:14.240
is inherent to programmers,
40:14.240 --> 40:16.920
is inherent to this kind of process?
40:16.920 --> 40:20.320
Or can it be improved?
40:20.320 --> 40:24.040
Can it be happy land where that tension
40:24.040 --> 40:26.320
is not quite so harsh?
40:26.320 --> 40:28.120
I don't think Stack Overflow
40:28.120 --> 40:30.040
can totally change the way it works.
40:30.040 --> 40:31.800
One thing they are working on, finally,
40:31.800 --> 40:35.200
is the ask page had not changed since 2011.
40:35.200 --> 40:36.760
I'm still kind of bitter about this
40:36.760 --> 40:39.360
because I feel like you have a Q&A system
40:39.360 --> 40:41.400
and one of the core pages in a Q&A system.
40:41.400 --> 40:42.240
Well, first of all, the question,
40:42.240 --> 40:43.760
all the answers and also the ask page,
40:43.760 --> 40:45.160
particularly when you're a new user
40:45.160 --> 40:46.520
or someone trying to ask a question,
40:46.520 --> 40:48.440
that's the point at which you need the most help.
40:48.440 --> 40:50.760
And we just didn't adapt with the times.
40:50.760 --> 40:52.360
But the good news is they're working on this
40:52.360 --> 40:53.240
from what I understand,
40:53.240 --> 40:55.200
and it's gonna be a more wizard based format.
40:55.200 --> 40:56.840
And you could envision a world
40:56.840 --> 40:58.440
where as part of this wizard based program,
40:58.440 --> 40:59.360
when you're asking a question,
40:59.360 --> 41:01.120
so okay, come up with a good title.
41:01.120 --> 41:02.560
What are good words to put in the title?
41:02.560 --> 41:04.280
One word that's not good to put in the title
41:04.280 --> 41:05.880
is problem, for example.
41:05.880 --> 41:06.720
I have a problem.
41:06.720 --> 41:07.800
Oh, you have a problem.
41:07.800 --> 41:10.000
Okay, a problem, that's great, right?
41:10.000 --> 41:11.240
Like, you need specifics, right?
41:11.240 --> 41:12.200
Like, so it's trying to help you
41:12.200 --> 41:14.240
make a good question title, for example.
41:14.240 --> 41:17.280
That step will be broken out, all that stuff.
41:17.280 --> 41:19.560
But one of those steps in that wizard of asking
41:19.560 --> 41:21.480
could say, hey, I'm a little nervous.
41:21.480 --> 41:22.680
You know, I've never done this before.
41:22.680 --> 41:26.080
Can you put me in a queue for like special mentoring, right?
41:26.080 --> 41:27.560
You could opt in to a special mentor.
41:27.560 --> 41:28.800
I think that would be fantastic.
41:28.800 --> 41:31.360
Like, I don't have any objection to that at all
41:31.360 --> 41:32.600
in terms of being an opt in system.
41:32.600 --> 41:33.880
Cause there are people that are like, you know,
41:33.880 --> 41:35.480
I just wanna help them.
41:35.480 --> 41:36.720
I wanna help a person no matter what.
41:36.720 --> 41:37.920
I wanna go above and beyond.
41:37.920 --> 41:40.160
I wanna spend like hours with this person.
41:40.160 --> 41:42.920
It depends what their goals are, right?
41:42.920 --> 41:43.760
It's a great idea.
41:43.760 --> 41:44.600
Who am I to judge, right?
41:44.600 --> 41:45.440
So that's fine.
41:45.440 --> 41:47.320
It's not precluded from happening.
41:47.320 --> 41:49.000
But there's a certain big city ethos
41:49.000 --> 41:51.080
that we started with of like, look, we're of New York City.
41:51.080 --> 41:52.800
You don't come to New York City and expect them to be,
41:52.800 --> 41:54.360
oh, welcome to the city, Joe.
41:54.360 --> 41:55.200
How's it going?
41:55.200 --> 41:56.040
Come on in.
41:56.040 --> 41:57.120
Let me show you around.
41:57.120 --> 41:59.360
That's not how New York City works, right?
41:59.360 --> 42:02.160
I mean, and you know, again, New York City
42:02.160 --> 42:03.080
is a reputation for being rude,
42:03.080 --> 42:04.080
which I actually don't think it is
42:04.080 --> 42:05.240
having been there fairly recently.
42:05.240 --> 42:06.080
It's not rude.
42:06.080 --> 42:07.200
People are just like going about their business, right?
42:07.200 --> 42:08.960
Like, look, I have things to do.
42:08.960 --> 42:11.800
I'm busy, I'm a busy professional, as are you.
42:11.800 --> 42:13.400
And since you're a busy professional,
42:13.400 --> 42:14.640
certainly when you ask a question,
42:14.640 --> 42:16.400
you're gonna ask the best possible question, right?
42:16.400 --> 42:18.200
Because you're a busy professional
42:18.200 --> 42:20.120
and you would not accept anything less
42:20.120 --> 42:21.400
than a very well went in question
42:21.400 --> 42:23.240
with a lot of detail about why you're doing it,
42:23.240 --> 42:25.040
what you're doing, what you researched,
42:25.040 --> 42:26.160
what you found, right?
42:26.160 --> 42:28.560
Cause you're a professional like me, right?
42:28.560 --> 42:30.960
And this rubs people sometimes the wrong way.
42:30.960 --> 42:32.280
And I don't think it's wrong to say,
42:32.280 --> 42:33.640
look, I don't want that experience.
42:33.640 --> 42:37.600
I want just a more chill place for beginners.
42:37.600 --> 42:39.160
And I still think Stack Overflow
42:39.160 --> 42:41.960
is not, was never designed for beginners, right?
42:41.960 --> 42:43.840
There's this misconception that, you know,
42:43.840 --> 42:45.160
even Joel says sometimes, oh yes,
42:45.160 --> 42:46.120
Stack Overflow for beginners.
42:46.120 --> 42:48.840
And I think if you're a prodigy, it can be.
42:48.840 --> 42:51.720
But that's not really representative, right?
42:51.720 --> 42:53.080
Like, I think as a beginner,
42:53.080 --> 42:55.120
you want a totally different set of tools.
42:55.120 --> 42:58.040
You want like live screen sharing, live chat.
42:58.040 --> 42:59.240
You want access to resources.
42:59.240 --> 43:01.360
You want a playground, like a playground
43:01.360 --> 43:03.280
you can experiment in and like test
43:03.280 --> 43:05.640
and all the stuff that we just don't give people
43:05.640 --> 43:07.760
because that was never really the audience
43:07.760 --> 43:09.280
that we were designing Stack Overflow for.
43:09.280 --> 43:10.680
That doesn't mean it's wrong.
43:10.680 --> 43:11.760
And I think it would be awesome
43:11.760 --> 43:13.240
if there was a site like that on the internet
43:13.240 --> 43:14.280
or if Stack Overflow said,
43:14.280 --> 43:15.800
hey, you know, we're going to start doing this.
43:15.800 --> 43:16.960
That's fine too.
43:16.960 --> 43:17.800
You know, I'm not there.
43:17.800 --> 43:18.640
I'm not making those decisions.
43:18.640 --> 43:21.040
But I do think the pressure or the tension
43:21.040 --> 43:23.560
that you described is there for people to be, look,
43:23.560 --> 43:24.440
I'm a little nervous
43:24.440 --> 43:26.360
cause I know I got to do my best work, right?
43:26.360 --> 43:28.000
The other one is something you talk about
43:28.000 --> 43:30.280
which is also really interesting to me
43:30.280 --> 43:32.040
is duplicate questions.
43:32.040 --> 43:37.040
Or it's a really difficult problem that you highlight.
43:38.080 --> 43:39.440
It's super hard.
43:39.440 --> 43:41.680
Like you could take one little topic
43:41.680 --> 43:44.920
and you could probably write 10, 20, 30 ways
43:44.920 --> 43:48.160
of asking about that topic and there will be all different.
43:48.160 --> 43:50.160
I don't know if there should be one page
43:50.160 --> 43:51.840
that answers all of it.
43:51.840 --> 43:56.680
Is there a way that Stack Overflow can help
43:57.840 --> 44:01.120
disambiguate, like separate these duplicate questions
44:01.120 --> 44:02.800
or connect them together?
44:02.800 --> 44:06.200
Or is it a totally hopeless, difficult, impossible task?
44:06.200 --> 44:08.440
I think it's a very, very hard computer science problem.
44:08.440 --> 44:10.280
And partly cause people are very good
44:10.280 --> 44:11.760
at using completely different words.
44:11.760 --> 44:13.160
It always amazed me on Stack Overflow,
44:13.160 --> 44:15.440
you'd have two questions that were functionally identical.
44:15.440 --> 44:17.520
And one question had like zero words in common
44:17.520 --> 44:18.360
with the other question.
44:18.360 --> 44:20.920
Like, oh my God, from a computer science perspective,
44:20.920 --> 44:22.880
how do you even begin to solve that?
44:22.880 --> 44:24.480
And it happens all the time.
44:24.480 --> 44:26.240
People are super good at this, right?
44:26.240 --> 44:28.320
Accidentally at asking the same thing
44:28.320 --> 44:31.200
in like 10, 20 different ways.
44:31.200 --> 44:32.280
And the other complexities,
44:32.280 --> 44:33.880
we want some of those duplicates to exist
44:33.880 --> 44:35.960
because if there's five versions with different words,
44:35.960 --> 44:37.280
have those five versions point
44:37.280 --> 44:39.560
to the one centralized answer, right?
44:39.560 --> 44:41.840
It's like, okay, this is duplicate, no worries.
44:41.840 --> 44:43.440
Here's the answer that you wanted over here
44:43.440 --> 44:47.960
on this, the prime example that we wanna have,
44:47.960 --> 44:50.120
rather than having 10 copies of the question
44:50.120 --> 44:50.960
and the answer.
44:50.960 --> 44:52.360
Because if you have 10 copies of the question answer,
44:52.360 --> 44:54.080
this also devalues the reputation system,
44:54.080 --> 44:56.840
which programmers hate, as I previously mentioned.
44:56.840 --> 44:58.280
You're getting reputation for an answer
44:58.280 --> 44:59.280
that somebody else already gave.
44:59.280 --> 45:00.400
It's like, well, it's an answer,
45:00.400 --> 45:02.160
but somebody else already gave that answer.
45:02.160 --> 45:04.520
So why are you getting reputation for the same answer
45:04.520 --> 45:06.440
as the other guy who gave it four years ago?
45:06.440 --> 45:07.720
People get offended by that, right?
45:07.720 --> 45:11.880
So the reputation system itself adds tension to the system.
45:11.880 --> 45:14.000
In that the people who have a lot of reputation
45:14.000 --> 45:17.560
become very incentivized to enforce the reputation system.
45:17.560 --> 45:18.880
And for the most part, this good.
45:18.880 --> 45:20.640
I know it sounds weird, but for most parts like,
45:20.640 --> 45:23.840
look, strict systems, I think to use Stack Overflow,
45:23.840 --> 45:24.800
you have to have the idea that, okay,
45:24.800 --> 45:26.600
strict systems ultimately work better.
45:26.600 --> 45:28.880
And I do think in programming, you're familiar
45:28.880 --> 45:31.280
with loose typing versus strict typing, right?
45:31.280 --> 45:32.960
The idea that you can declare a variable,
45:32.960 --> 45:34.000
not declare a variable, rather,
45:34.000 --> 45:35.080
just start using a variable and, okay,
45:35.080 --> 45:37.160
I see it's implicitly an integer, bam, awesome.
45:37.160 --> 45:38.120
Duck equals five.
45:38.120 --> 45:40.760
Well, duck is now an integer of five, right?
45:40.760 --> 45:42.320
And you're like, cool, awesome, simpler, right?
45:42.320 --> 45:43.800
Why would I want to worry about typing?
45:43.800 --> 45:45.800
And for a long time, like in the Ruby community,
45:45.800 --> 45:46.760
they're like, yeah, this is awesome.
45:46.760 --> 45:48.160
Like you just do a bunch of unit testing,
45:48.160 --> 45:50.920
which is testing your program's validity after the fact
45:50.920 --> 45:54.160
to catch any bugs that strict typing of variables
45:54.160 --> 45:55.000
would have caught.
45:55.000 --> 45:56.680
And now you have this thing called TypeScript
45:56.680 --> 45:59.560
from Microsoft, from the guy who built C Sharp Anders,
45:59.560 --> 46:02.680
who's one of the greatest minds in software development,
46:02.680 --> 46:04.200
right, like in terms of language design,
46:04.200 --> 46:06.080
and says, no, no, no, we want to bolt on
46:06.080 --> 46:07.320
a strict type system to JavaScript
46:07.320 --> 46:08.160
because it makes things better.
46:08.160 --> 46:10.000
And now everybody's like, oh my God,
46:10.000 --> 46:13.200
we deployed TypeScript and found 50 lane bugs
46:13.200 --> 46:14.480
that we didn't know about, right?
46:14.480 --> 46:15.800
Like this is super common.
46:15.800 --> 46:19.880
So I think there is a truth in programming
46:19.880 --> 46:22.200
that strictness, it's not the goal.
46:22.200 --> 46:23.480
We're not saying be super strict
46:23.480 --> 46:25.680
because strictness is correct.
46:25.680 --> 46:28.600
No, it's no, no, strictness produces better results.
46:28.600 --> 46:29.840
That's what I'm saying, right?
46:29.840 --> 46:31.720
So strict typing of variables,
46:31.720 --> 46:33.880
I would say you almost universally have consensus now
46:33.880 --> 46:35.400
is basically correct.
46:35.400 --> 46:37.680
Should be that way in every language, right?
46:37.680 --> 46:39.480
Duck equals five should generate an error
46:39.480 --> 46:40.680
because no, you didn't declare,
46:40.680 --> 46:42.920
you didn't tell me that duck was an integer, right?
46:42.920 --> 46:43.840
That's a bug, right?
46:43.840 --> 46:45.720
Or maybe you mistyped, you typed duck, right?
46:45.720 --> 46:46.560
Instead of duck, right?
46:46.560 --> 46:48.800
You never know, this happens all the time, right?
46:48.800 --> 46:51.600
So with that in mind, I will say that the strictness
46:51.600 --> 46:52.680
of the system is correct.
46:52.680 --> 46:55.040
Now, that doesn't mean cruel, that doesn't mean mean,
46:55.040 --> 46:57.800
that doesn't mean angry, it just means strict, okay?
46:57.800 --> 46:59.280
So I think where there's misunderstanding
46:59.280 --> 47:00.720
is in people get cranky, right?
47:00.720 --> 47:02.280
Like another question you asked is like,
47:02.280 --> 47:05.120
why are programmers kind of mean sometimes?
47:05.120 --> 47:07.360
Well, who do programmers work with all day long?
47:07.360 --> 47:10.040
So I have a theory that if you're at a job
47:10.040 --> 47:12.400
and you work with assholes all day long,
47:12.400 --> 47:14.320
what do you eventually become?
47:14.320 --> 47:15.160
An asshole.
47:15.160 --> 47:17.040
An asshole, and what is the computer
47:17.040 --> 47:19.240
except the world's biggest asshole?
47:19.240 --> 47:21.800
Because the computer has no time for your bullshit.
47:21.800 --> 47:24.000
The computer, the minute you make a mistake,
47:24.000 --> 47:25.240
everything is crashing down, right?
47:25.240 --> 47:28.080
One semicolon has crashed space missions, right?
47:28.080 --> 47:29.000
So that's normal.
47:29.000 --> 47:30.600
So you begin to internalize that.
47:30.600 --> 47:34.560
You begin to think, oh, my coworker, the computer,
47:34.560 --> 47:37.440
is super strict and kind of a jerk about everything.
47:37.440 --> 47:39.960
So that's kind of how I'm gonna be
47:39.960 --> 47:41.800
because I work with this computer
47:41.800 --> 47:44.520
and I have to exceed to its terms on everything.
47:44.520 --> 47:46.360
So therefore, you start to absorb that
47:46.360 --> 47:48.520
and you start to think, oh, well, being really strict
47:48.520 --> 47:50.360
arbitrarily is really good.
47:50.360 --> 47:54.120
An error code 56249 is a completely good error message
47:54.120 --> 47:56.080
because that's what the computer gave me, right?
47:56.080 --> 47:59.320
So you kind of forget to be a person at some level.
47:59.320 --> 48:01.400
And you know how they say great detectives
48:01.400 --> 48:04.200
internalize criminals and kind of arc criminals themselves
48:04.200 --> 48:07.080
like this trope of the master detective is good
48:07.080 --> 48:08.680
because you can think like the criminal.
48:08.680 --> 48:10.400
Well, I do think that's true of programmers.
48:10.400 --> 48:12.640
Really good programmers think like the computer
48:12.640 --> 48:14.200
because that's their job.
48:14.200 --> 48:17.080
But if you internalize it too much, you become the computer
48:17.080 --> 48:19.480
and you kind of become a jerk to everybody
48:19.480 --> 48:21.680
because that's what you've internalized.
48:21.680 --> 48:22.640
You're almost not a jerk,
48:22.640 --> 48:25.600
but you have no patience for lack of strictness,
48:25.600 --> 48:26.440
as you said.
48:26.440 --> 48:27.800
It's not out of a sense of meanness.
48:27.800 --> 48:28.640
It's accidental,
48:28.640 --> 48:30.360
but I do believe it's an occupational hazard
48:30.360 --> 48:32.440
or being a programmer is you start to behave
48:32.440 --> 48:33.680
like the computer.
48:33.680 --> 48:35.840
You're very unforgiving, you're very terse,
48:35.840 --> 48:38.400
you're very, oh, wrong, incorrect, move on.
48:38.400 --> 48:39.720
It's like, well, can you help me?
48:39.720 --> 48:41.200
Like what could I do to fix?
48:41.200 --> 48:43.960
No, wrong, next question, right?
48:43.960 --> 48:46.400
Like that's normal for the computer, right?
48:46.400 --> 48:47.760
Just fail next, right?
48:47.760 --> 48:51.240
Like, I don't know if you remember in Saturday Night Live,
48:51.240 --> 48:53.080
like in the nineties, they had this character
48:53.080 --> 48:56.720
who was an IT guy, the move guy, move.
48:56.720 --> 48:58.120
Was that Jimmy Fallon?
48:58.120 --> 49:01.160
No, no, who played him?
49:01.160 --> 49:02.640
Okay, yeah, I remember move.
49:02.640 --> 49:04.200
Right, he had no patience for him.
49:04.200 --> 49:05.720
Might have been mad TV actually.
49:05.720 --> 49:06.680
Oh, was it mad TV?
49:06.680 --> 49:07.600
Might have been, might have been.
49:07.600 --> 49:10.960
But anyway, that's always been the perception, right?
49:10.960 --> 49:12.560
You start to behave like the computer.
49:12.560 --> 49:14.920
It's like, oh, you're wrong out of the way, you know?
49:14.920 --> 49:17.560
You've written so many blog posts about programming,
49:17.560 --> 49:22.000
about programs, programming, programmers.
49:22.640 --> 49:25.640
What do you think makes a good,
49:25.640 --> 49:29.320
let's start with, what makes a good solo programmer?
49:29.320 --> 49:31.560
Well, I don't think you should be a solo programmer.
49:31.560 --> 49:33.320
I think to be a good solo programmer,
49:33.320 --> 49:35.560
it's kind of like what I talked about, well, not on mic,
49:35.560 --> 49:38.000
but one of the things, John Carmack,
49:38.000 --> 49:40.080
one of the best points he makes in the book,
49:40.080 --> 49:42.120
Masters of Doom, which is a fantastic book,
49:42.120 --> 49:43.680
and anybody listening to this who hasn't read it,
49:43.680 --> 49:45.800
please read it, it's such a great book,
49:45.800 --> 49:48.520
is that at the time they were working on stuff
49:48.520 --> 49:50.400
like Wolfenstein and Doom,
49:50.400 --> 49:52.800
like they didn't have the resources that we have today.
49:52.800 --> 49:53.960
They didn't have Stack Overflow,
49:53.960 --> 49:54.880
they didn't have Wikipedia,
49:54.880 --> 49:56.680
they didn't have like discourse forums,
49:56.680 --> 49:58.720
they didn't have places to go
49:58.720 --> 50:00.200
to get people to help them, right?
50:00.200 --> 50:01.920
They had to work on their own,
50:01.920 --> 50:04.720
and that's why it took a genius like Carmack to do this stuff,
50:04.720 --> 50:05.560
because you had to be a genius
50:05.560 --> 50:07.160
to invent from First Pensibles,
50:07.160 --> 50:08.560
a lot of the stuff he was like,
50:08.560 --> 50:11.040
the hacks he was coming up with were genius, right?
50:11.040 --> 50:12.520
Genius level stuff, but you don't need
50:12.520 --> 50:13.840
to be a genius anymore,
50:13.840 --> 50:15.560
and that means not working by yourself.
50:15.560 --> 50:17.400
You have to be good at researching stuff online,
50:17.400 --> 50:19.240
you have to be good at asking questions,
50:19.240 --> 50:21.040
really good questions that are really well researched,
50:21.040 --> 50:22.400
which implies, oh, I went out and researched
50:22.400 --> 50:24.120
for three hours before I wrote these questions,
50:24.120 --> 50:25.680
like that's what you should be doing,
50:25.680 --> 50:28.160
because that's what's gonna make you good, right?
50:28.160 --> 50:29.920
To me, this is the big difference between programming
50:29.920 --> 50:32.160
in like the 80s versus programming today,
50:32.160 --> 50:34.920
is like you kind of had to be by yourself back then,
50:34.920 --> 50:36.520
like where would you go for answers?
50:36.520 --> 50:37.920
I remember in the early days
50:37.920 --> 50:41.120
when I was learning Visual Basic for Windows,
50:41.120 --> 50:44.600
like I would call the Microsoft Helpline on the phone
50:44.600 --> 50:45.680
when I had like program,
50:45.680 --> 50:47.040
because I was like, I don't know what to do.
50:47.040 --> 50:48.560
So I would like go and call,
50:48.560 --> 50:49.640
and they had these huge phone banks,
50:49.640 --> 50:51.280
and like, can you imagine how alien that is now?
50:51.280 --> 50:52.360
Like who would do that, right?
50:52.360 --> 50:53.520
Like that's crazy.
50:53.520 --> 50:57.000
So there was just nowhere else to go when you got stuck,
50:57.000 --> 50:57.840
right?
50:57.840 --> 50:59.520
Like I had the books that came with it,
50:59.520 --> 51:01.320
I read those, studied those religiously.
51:01.320 --> 51:03.680
I just saw a post from Steve Sinovsky that said,
51:03.680 --> 51:08.680
this C++ version seven came with like 10,000 pages
51:08.880 --> 51:10.600
of written material,
51:10.600 --> 51:13.600
because where else were you gonna figure that stuff out?
51:13.600 --> 51:14.880
You didn't go to the library?
51:14.880 --> 51:15.800
I mean, you didn't have Wikipedia,
51:15.800 --> 51:17.880
you didn't have Reddit,
51:17.880 --> 51:20.760
you didn't have anywhere to go to answer these questions.
51:20.760 --> 51:24.200
So you've talked about through the years,
51:24.200 --> 51:25.560
basically not having an ego
51:25.560 --> 51:28.840
and not thinking that you're the best programmer
51:28.840 --> 51:29.840
in the world,
51:29.840 --> 51:34.120
and so always kind of just looking to improve,
51:34.120 --> 51:36.520
to become a better programmer than you were yesterday.
51:36.520 --> 51:39.240
So how have you changed as a programmer
51:39.240 --> 51:43.320
and as a thinker designer around programming
51:43.320 --> 51:47.360
over the past 15 years,
51:47.360 --> 51:49.000
really of being a public figure?
51:49.000 --> 51:50.880
I would say the big insight that I had
51:50.880 --> 51:52.680
is eventually as a programmer,
51:52.680 --> 51:54.760
you have to kind of stop writing code to be effective,
51:54.760 --> 51:57.000
which is kind of disturbing,
51:57.000 --> 51:58.280
because you really love it.
51:58.280 --> 52:02.160
But you realize like being effective at programming
52:02.160 --> 52:05.160
in the general sense doesn't mean writing code.
52:05.160 --> 52:06.620
And a lot of times you can be much more successful
52:06.620 --> 52:08.840
by not writing code and writing code
52:08.840 --> 52:10.400
in terms of just solving the problems you have,
52:10.400 --> 52:12.400
essentially hiring people that are really good
52:12.400 --> 52:13.800
and like setting them free
52:13.800 --> 52:16.240
and like giving them basic direction, right?
52:16.240 --> 52:18.040
Like on strategy and stuff,
52:18.040 --> 52:19.680
because a lot of the problems you encounter
52:19.680 --> 52:22.400
aren't necessarily solved to like really gnarly code,
52:22.400 --> 52:25.080
they're solved by conceptual solutions,
52:25.080 --> 52:26.560
which can then be turned into code,
52:26.560 --> 52:28.680
but are you even solving the right problem?
52:28.680 --> 52:31.920
I mean, so I would say for me,
52:31.920 --> 52:35.320
the main insight I have is to succeed as a programmer,
52:35.320 --> 52:37.560
you eventually kind of stop writing code.
52:37.560 --> 52:38.720
That's gonna sound discouraging,
52:38.720 --> 52:40.760
probably to people hearing, but I don't mean it that way.
52:40.760 --> 52:43.320
What I mean is that you're coding in a higher level language.
52:43.320 --> 52:44.280
Eventually like, okay,
52:44.280 --> 52:45.600
so we're coding in assembly language, right?
52:45.600 --> 52:46.640
That's the beginning, right?
52:46.640 --> 52:48.960
You're hard coded to the architecture.
52:48.960 --> 52:49.800
Then you have stuff like C,
52:49.800 --> 52:51.000
where it's like, wow, we can abstract
52:51.000 --> 52:52.600
across the architecture, you can write code.
52:52.600 --> 52:56.320
I can then compile that code for ARM or whatever,
52:56.320 --> 52:58.280
XA6 or whatever else is out there.
52:58.280 --> 53:00.120
And then even higher level NAT, right?
53:00.120 --> 53:01.920
Like you're looking at like Python, Ruby,
53:01.920 --> 53:03.000
interpreted languages.
53:03.000 --> 53:05.000
And then to me as a programmer, like, okay,
53:05.000 --> 53:06.000
I wanna go even higher.
53:06.000 --> 53:07.160
I wanna go higher than that.
53:07.160 --> 53:08.560
How do I abstract higher than language?
53:08.560 --> 53:11.120
It's like, well, you abstract in spoken language
53:11.120 --> 53:12.120
and written language, right?
53:12.120 --> 53:14.600
Like you're sort of inspiring people to get things done,
53:14.600 --> 53:16.080
giving them guidance, like, what if we did this?
53:16.080 --> 53:17.480
What if we did this?
53:17.480 --> 53:19.800
You're writing in the highest level language that there is,
53:19.800 --> 53:21.360
which is for me English, right?
53:21.360 --> 53:23.040
Whatever your spoken language is.
53:23.040 --> 53:25.840
So it's all about being effective, right?
53:25.840 --> 53:29.360
And I think Patrick McKenzie,
53:29.360 --> 53:32.000
patio 11 on Hacker News and Works at Stripe,
53:32.000 --> 53:33.600
has a great post about this,
53:33.600 --> 53:35.880
of how calling yourself a programmer
53:35.880 --> 53:38.600
is a career limiting move at some level.
53:38.600 --> 53:39.760
Once you get far enough from your career.
53:39.760 --> 53:40.800
And I really believe that.
53:40.800 --> 53:43.080
And again, I apologize, this is sound discouraging.
53:43.080 --> 53:45.280
I don't mean it to be, but he's so right.
53:45.280 --> 53:47.760
Because all the stuff that goes on around the code,
53:47.760 --> 53:49.800
like the people, like that's another thing.
53:49.800 --> 53:51.400
If you look at my early blog interviews, it was about,
53:51.400 --> 53:55.280
wow, programming is about people more than it's about code,
53:55.280 --> 53:56.720
which doesn't really make sense, right?
53:56.720 --> 53:59.200
But it's about, can these people even get along together?
53:59.200 --> 54:00.880
Can they understand each other?
54:00.880 --> 54:03.000
Can you even explain to me what it is you're working on?
54:03.000 --> 54:04.560
Are you solving the right problem?
54:04.560 --> 54:06.600
People wear it, right? Another classic programming book,
54:06.600 --> 54:08.520
which again, up there with code complete,
54:08.520 --> 54:09.960
please read people wear.
54:09.960 --> 54:11.600
It's that software is people, right?
54:11.600 --> 54:13.520
People are the software first and foremost.
54:13.520 --> 54:15.680
So a lot of the skills that I was working on
54:15.680 --> 54:17.600
early in the blog were about
54:17.600 --> 54:19.400
figuring out the people parts of programming,
54:19.400 --> 54:20.680
which was the harder parts.
54:20.680 --> 54:21.680
The hard part of programming,
54:21.680 --> 54:23.080
once you get to certain skill level programming,
54:23.080 --> 54:25.240
you can pretty much solve any reasonable problem
54:25.240 --> 54:26.320
that's put in front of you.
54:26.320 --> 54:28.120
You're not writing algorithms from scratch, right?
54:28.120 --> 54:29.280
That just doesn't happen.
54:29.280 --> 54:30.560
So any sort of reasonable problem
54:30.560 --> 54:32.280
put in front of you, you're gonna be able to solve.
54:32.280 --> 54:33.600
But what you can't solve is,
54:33.600 --> 54:36.920
our manager is a total jerk.
54:36.920 --> 54:38.600
You cannot solve that with code.
54:38.600 --> 54:40.640
That is not a code solvable problem.
54:40.640 --> 54:43.680
And yet that will cripple you way more than,
54:43.680 --> 54:45.800
oh, we had to use this stupid framework I don't like,
54:45.800 --> 54:49.320
or Sam keeps writing bad code that I hate,
54:49.320 --> 54:52.160
or Dave is off there in the wilderness,
54:52.160 --> 54:53.280
writing God knows what, right?
54:53.280 --> 54:54.120
These are not your problems.
54:54.120 --> 54:56.800
Your problem is your manager or a coworker
54:56.800 --> 54:58.760
is so toxic to everybody else in your team
54:58.760 --> 55:00.160
that nobody can get anything done
55:00.160 --> 55:02.320
because everybody's so stressed out and freaked out, right?
55:02.320 --> 55:04.480
These are the problems that you have to attack.
55:04.480 --> 55:05.320
Absolutely.
55:05.320 --> 55:07.360
And so as you go to these higher level abstractions
55:07.360 --> 55:09.080
as you've developed as a programmer
55:09.080 --> 55:10.600
to higher, higher level abstractions
55:10.600 --> 55:12.240
and go into natural language,
55:12.240 --> 55:15.680
you're also the guy who kind of preached,
55:15.680 --> 55:18.640
building it, diving in and doing it,
55:18.640 --> 55:21.760
and like learn by doing.
55:21.760 --> 55:22.840
Yes.
55:22.840 --> 55:27.840
Do you worry that as you get to higher,
55:28.480 --> 55:30.040
higher level abstractions,
55:30.040 --> 55:35.040
that you lose track of the lower level of just building
55:35.440 --> 55:37.200
is like, do you worry about that?
55:37.200 --> 55:41.600
Not maybe now, but 10 years from now, 20 years from now?
55:41.600 --> 55:42.520
Well, no.
55:42.520 --> 55:43.840
I mean, there is always that paranoia
55:43.840 --> 55:45.400
and oh gosh, I don't feel it's valuable
55:45.400 --> 55:46.240
since I'm not writing code.
55:46.240 --> 55:48.280
But for me, when we started the discourse project,
55:48.280 --> 55:50.280
it was Ruby, which I didn't really know Ruby.
55:50.280 --> 55:51.600
I mean, as you pointed out,
55:51.600 --> 55:53.840
and this is another valuable abstract from Stack Overflow,
55:53.840 --> 55:56.040
you can be super proficient at, for example, C Sharp,
55:56.040 --> 55:56.880
which I was working in,
55:56.880 --> 55:57.880
that's what we built Stack Overflow
55:57.880 --> 55:59.240
and then still is written in.
55:59.240 --> 56:01.480
And then switched to Ruby and you're a newbie again, right?
56:01.480 --> 56:03.200
But you have the framework.
56:03.200 --> 56:04.400
I know what a for loop is.
56:04.400 --> 56:05.760
I know what recursion is.
56:05.760 --> 56:09.600
I know what a stack trace is, right?
56:09.600 --> 56:11.360
Like I have all the fundamental concepts
56:11.360 --> 56:12.760
to be a programmer, I just don't know Ruby.
56:12.760 --> 56:14.760
So I'm still on a higher level.
56:14.760 --> 56:16.560
I'm not like a beginner beginner, like you're saying.
56:16.560 --> 56:18.560
I'm just like, I need to apply my programming concepts
56:18.560 --> 56:20.280
I already know to Ruby.
56:20.280 --> 56:23.160
Well, so there's a question that's really interesting.
56:23.160 --> 56:26.840
So looking at Ruby, how do you go about learning enough
56:26.840 --> 56:28.560
that your intuition can be applied?
56:28.560 --> 56:30.320
Well, that's the thing, that's what I was trying to get to
56:30.320 --> 56:31.160
is like what I realized,
56:31.160 --> 56:32.360
particularly when I started with just me and Robin,
56:32.360 --> 56:34.800
I realized if I bother Robin,
56:34.800 --> 56:37.080
I am now costing us productivity, right?
56:37.080 --> 56:39.640
Every time I go to Robin, rather than building
56:39.640 --> 56:43.760
the our first alpha version of discourse,
56:43.760 --> 56:46.040
he's now answering my stupid questions about Ruby.
56:46.040 --> 56:47.600
Is that a good use of his time?
56:47.600 --> 56:49.160
Is that a good use of my time?
56:49.160 --> 56:52.320
The answer to both of those was resoundingly no, right?
56:52.320 --> 56:54.320
Like we were getting to an alpha
56:54.320 --> 56:55.320
and it was pretty much like, okay,
56:55.320 --> 56:56.720
we'll hire more programmers, right?
56:56.720 --> 56:59.000
Like we eventually hired Neil
56:59.000 --> 57:01.760
and then eventually Sam, who came in as a cofounder.
57:02.880 --> 57:05.040
Actually it was Sam first, then Neil later.
57:05.040 --> 57:06.280
But the answer to the problem
57:06.280 --> 57:07.560
is just hire other competent programmers.
57:07.560 --> 57:10.360
It's not like teach, now I shall pull myself up
57:10.360 --> 57:12.360
by my bootstraps and learn Ruby.
57:12.360 --> 57:16.080
But at some point writing code becomes a liability to you
57:16.080 --> 57:17.400
in terms of getting things done.
57:17.400 --> 57:19.400
There's so many other things that go on in the project,
57:19.400 --> 57:20.440
like building the prototype.
57:20.440 --> 57:22.400
Like you mentioned like, well, how do you,
57:22.400 --> 57:23.240
if you're not writing code,
57:23.240 --> 57:25.720
how does everybody keep focus on like what are we building?
57:25.720 --> 57:28.160
Well, first basic mockups and research, right?
57:28.160 --> 57:30.160
Like what do we even want to build?
57:30.160 --> 57:31.240
There's a little bit of that that goes on.
57:31.240 --> 57:32.800
But then very quickly you get to the prototype stage.
57:32.800 --> 57:33.800
Like build a prototype,
57:33.800 --> 57:35.960
let's iterate on the prototype really, really rapidly.
57:35.960 --> 57:36.880
And that's what we do with discourse.
57:36.880 --> 57:38.960
And that's what we demoed to get our seed funding
57:38.960 --> 57:42.360
for discourse was the alpha version of discourse
57:42.360 --> 57:43.840
that we had running and ready to go.
57:43.840 --> 57:45.160
And it was very, it was bad.
57:45.160 --> 57:47.400
I mean, it was, I'll just tell you it was bad.
57:47.400 --> 57:48.440
I have, we have screenshots of it
57:48.440 --> 57:50.560
and I'm just like embarrassed to look at it now.
57:50.560 --> 57:51.400
But it was the prototype.
57:51.400 --> 57:52.640
We were figuring out like what's working,
57:52.640 --> 57:53.560
what's not working.
57:53.560 --> 57:56.360
Cause there's such a broad gap between,
57:56.360 --> 57:59.440
between the way you think things will work in your mind
57:59.440 --> 58:01.160
or even on paper and the way they work
58:01.160 --> 58:03.080
once you sit and live in the software,
58:03.080 --> 58:05.600
like actually spend time living and breathing on software
58:05.600 --> 58:06.720
so different.
58:06.720 --> 58:10.800
So my philosophy is get to a prototype
58:10.800 --> 58:12.360
and then what you're really optimizing
58:12.360 --> 58:13.240
for a speed of iteration,
58:13.240 --> 58:14.800
like how you can turn the crank,
58:14.800 --> 58:16.120
how quickly can we iterate?
58:16.120 --> 58:18.000
That's the absolutely critical metric
58:18.000 --> 58:19.000
of any software project.
58:19.000 --> 58:20.920
And I had a tweet recently that people liked
58:20.920 --> 58:23.120
and I totally, this is so fundamental to what I do,
58:23.120 --> 58:25.960
is like, if you want to measure the core competency
58:25.960 --> 58:27.600
of any software tech company,
58:27.600 --> 58:29.120
it's the speed at which somebody can say,
58:29.120 --> 58:30.760
hey, we really need this word
58:30.760 --> 58:32.240
and the product changed to this word, right?
58:32.240 --> 58:34.240
Because it will be more clear to the users,
58:34.240 --> 58:36.400
like what, like instead of respond, it's reply or something.
58:36.400 --> 58:39.760
But there's some, from the conception of that idea
58:39.760 --> 58:42.160
to how quickly that single word can be changed
58:42.160 --> 58:43.400
in your software and rolled out to users,
58:43.400 --> 58:44.880
that is your life cycle.
58:44.880 --> 58:47.240
That's your health, your, your heartbeat.
58:47.240 --> 58:50.000
If your heartbeat is like super slow,
58:50.000 --> 58:51.600
you're basically dead.
58:51.600 --> 58:54.160
No, seriously, like if it takes two weeks
58:54.160 --> 58:56.240
or even a month to get that single word changed,
58:56.240 --> 58:57.720
that was like, oh my God, this is a great idea.
58:57.720 --> 58:59.000
That word is so much clearer.
58:59.000 --> 59:00.040
I'm talking about like a super,
59:00.040 --> 59:01.640
like everybody's on board for this change.
59:01.640 --> 59:03.440
It's not like let's just change a word because we're bored.
59:03.440 --> 59:05.120
It's like, this is an awesome change.
59:05.120 --> 59:07.000
And then it takes a month to roll out.
59:07.000 --> 59:07.880
It's like, well, you're dead.
59:07.880 --> 59:09.160
Like you can't iterate.
59:09.160 --> 59:11.640
You can't, how you can do anything, right?
59:11.640 --> 59:13.680
Like, so anyway, about the heartbeat,
59:13.680 --> 59:15.960
it's like get the prototype and then iterate on it.
59:15.960 --> 59:18.960
That's what I view as like the central tenet
59:18.960 --> 59:20.440
of some modern software development.
59:20.440 --> 59:22.000
That's fascinating that you put it that way.
59:22.000 --> 59:24.440
It's actually, so I work in, I build autonomous vehicles
59:24.440 --> 59:28.160
and when you look at what maybe compare Tesla
59:28.160 --> 59:31.400
to most other automakers, the psych,
59:31.400 --> 59:35.680
the whatever the heartbeat for Tesla is literally days.
59:35.680 --> 59:38.120
Now, in terms of they can over the air
59:38.120 --> 59:41.160
deploy software updates to all their vehicles,
59:41.160 --> 59:44.760
which is markedly different than every other automaker,
59:44.760 --> 59:49.080
which takes years to update a piece of software.
59:49.080 --> 59:52.120
And so, and that's reflected in everything
59:52.120 --> 59:55.520
that's the final product that's reflected
59:55.520 --> 59:58.200
in really how slowly they adapt to the times.
59:58.200 --> 59:59.920
And to be clear, I'm not saying being a hummingbird
59:59.920 --> 1:00:00.760
is the goal either.
1:00:00.760 --> 1:00:02.400
It's like, you don't want a heartbeat that's like so fast.
1:00:02.400 --> 1:00:04.560
It's like, you're just freaking out.
1:00:04.560 --> 1:00:05.680
But like, it is a measure of health.
1:00:05.680 --> 1:00:07.280
You should have a healthy heartbeat.
1:00:07.280 --> 1:00:09.560
It's up to, for people listening to decide what that means,
1:00:09.560 --> 1:00:10.840
but it has to be healthy.
1:00:10.840 --> 1:00:11.680
It has to be reasonable
1:00:11.680 --> 1:00:13.360
because otherwise you're just gonna be frustrated
1:00:13.360 --> 1:00:15.120
because like that's how you build software.
1:00:15.120 --> 1:00:17.720
You make mistakes, you roll it out, you live with it.
1:00:17.720 --> 1:00:18.760
You see what it feels like and say,
1:00:18.760 --> 1:00:19.920
oh God, that was a terrible idea.
1:00:19.920 --> 1:00:22.720
Oh my gosh, this could be even better if we did why, right?
1:00:22.720 --> 1:00:25.000
You turn the crank and then the more you do that,
1:00:25.000 --> 1:00:27.520
the faster you get ahead of your competitors ultimately
1:00:27.520 --> 1:00:29.720
because it's rate of change, right?
1:00:29.720 --> 1:00:30.920
Delta V, right?
1:00:30.920 --> 1:00:32.680
How fast are you moving?
1:00:32.680 --> 1:00:34.760
Well, within a year, you're gonna be miles away
1:00:34.760 --> 1:00:36.560
by the time they catch up with you, right?
1:00:36.560 --> 1:00:37.920
Like that's the way it works.
1:00:37.920 --> 1:00:40.760
And plus users, like as a software developer,
1:00:40.760 --> 1:00:42.800
I love software that's constantly changing
1:00:42.800 --> 1:00:45.600
because I don't understand people who get super pissed off
1:00:45.600 --> 1:00:47.200
when like, oh, they changed the software on me.
1:00:47.200 --> 1:00:48.040
How dare they?
1:00:48.040 --> 1:00:50.600
Yes, change the software, change it all the time, man.
1:00:50.600 --> 1:00:52.520
That's what makes this stuff great
1:00:52.520 --> 1:00:55.160
is that it can be changed so rapidly
1:00:55.160 --> 1:00:58.240
and become something that is greater than it is now.
1:00:58.240 --> 1:01:00.960
Now granted, there's some changes that suck, I admit.
1:01:00.960 --> 1:01:02.040
I've seen it many times.
1:01:02.040 --> 1:01:04.640
But in general, it's like, that's what makes software cool,
1:01:04.640 --> 1:01:06.280
right, is that it is so malleable.
1:01:06.280 --> 1:01:08.240
Like fighting that is like weird to me
1:01:08.240 --> 1:01:09.880
because it's like, well, you're fighting
1:01:09.880 --> 1:01:11.840
the essence of the thing that you're building.
1:01:11.840 --> 1:01:12.800
Like that doesn't make sense.
1:01:12.800 --> 1:01:15.240
You wanna really embrace that, not to be a hummingbird,
1:01:15.240 --> 1:01:18.120
but like embrace it to a healthy cycle of your heartbeat, right?
1:01:18.120 --> 1:01:20.880
So you talk about that people really don't change.
1:01:20.880 --> 1:01:23.840
It's true, that's why probably a lot of the stuff
1:01:23.840 --> 1:01:27.440
you write about in your blog probably will remain true.
1:01:27.440 --> 1:01:29.200
Well, there's a flip side of the coin, people don't change.
1:01:29.200 --> 1:01:31.440
So investing and understanding people
1:01:31.440 --> 1:01:34.640
is like learning Unix in 1970
1:01:34.640 --> 1:01:36.840
because nothing has changed, right?
1:01:36.840 --> 1:01:38.680
Like all those things you've learned about people
1:01:38.680 --> 1:01:40.480
will still be valid 34 years from now.
1:01:40.480 --> 1:01:43.000
Whereas if you learn the latest JavaScript framework,
1:01:43.000 --> 1:01:45.040
that's gonna be good for like two years, right?
1:01:45.040 --> 1:01:46.200
Exactly.
1:01:46.200 --> 1:01:49.640
So, but if you look at the future of programming,
1:01:49.640 --> 1:01:51.000
so there's a people component,
1:01:51.000 --> 1:01:54.200
but there's also the technology itself.
1:01:54.200 --> 1:01:57.160
Do you, what do you see as the future of programming?
1:01:57.160 --> 1:01:59.160
Will it change significantly?
1:01:59.160 --> 1:02:00.760
Or as far as you can tell,
1:02:01.680 --> 1:02:05.800
people are ultimately programming and so it will not,
1:02:05.800 --> 1:02:08.520
it's not something that you foresee changing
1:02:08.520 --> 1:02:09.840
in any fundamental way.
1:02:09.840 --> 1:02:14.240
Well, you gotta go look back on sort of the basics of programming.
1:02:14.240 --> 1:02:15.480
And one of the things that always shocked me
1:02:15.480 --> 1:02:16.320
is like source control.
1:02:16.320 --> 1:02:18.480
Like I didn't learn anything about source control.
1:02:18.480 --> 1:02:22.360
And I graduated from college in 1992,
1:02:22.360 --> 1:02:24.160
but I remember hearing from people
1:02:24.160 --> 1:02:26.560
like as late as like 1998, 1999,
1:02:26.560 --> 1:02:28.880
like even maybe today they're not learning source control.
1:02:28.880 --> 1:02:31.520
And to me it's like, well, how can you not learn source control?
1:02:31.520 --> 1:02:34.560
That is so fundamental to working with other programmers,
1:02:34.560 --> 1:02:35.920
working in a way that you don't lose your work.
1:02:35.920 --> 1:02:38.120
Like just basic soft, the bed,
1:02:38.120 --> 1:02:41.160
literal bedrock software development is source control.
1:02:41.160 --> 1:02:42.760
Now you compare today like GitHub, right?
1:02:42.760 --> 1:02:44.320
Like Microsoft brought GitHub, which I think was
1:02:44.320 --> 1:02:46.920
an incredibly smart acquisition move on their part.
1:02:46.920 --> 1:02:50.040
Now they have anybody who wants like reasonable source controls
1:02:50.040 --> 1:02:52.360
and go sign up on GitHub, it's all set up for you, right?
1:02:52.360 --> 1:02:55.280
There's tons of walkthroughs, tons of tutorials.
1:02:55.280 --> 1:02:57.680
So from the concept of like has programming advanced
1:02:57.680 --> 1:03:00.280
from say 1999, it's like, well, hell, we have GitHub.
1:03:00.280 --> 1:03:01.560
I mean, my God, yes, right?
1:03:01.560 --> 1:03:04.800
Like it's massively advanced over what it was.
1:03:04.800 --> 1:03:08.840
Now as to whether programming is significantly different,
1:03:08.840 --> 1:03:11.560
I'm gonna say no, but I think the baseline of like
1:03:11.560 --> 1:03:14.480
what we view as like fundamentals
1:03:14.480 --> 1:03:16.480
will continue to go up and actually get better.
1:03:16.480 --> 1:03:17.480
Like source control, for example,
1:03:17.480 --> 1:03:19.040
that's one of the fundamentals that has gotten,
1:03:19.040 --> 1:03:22.000
I mean, hundreds of orders of magnitude better
1:03:22.000 --> 1:03:23.760
than it was 10, 20 years ago, right?
1:03:23.760 --> 1:03:24.920
So those are the fundamentals.
1:03:24.920 --> 1:03:28.240
Let me introduce two things that maybe you can comment on.
1:03:28.240 --> 1:03:31.000
So one is mobile phones.
1:03:31.000 --> 1:03:36.000
So that could fundamentally transform what programming is,
1:03:37.080 --> 1:03:39.080
or maybe not, maybe you can comment on that.
1:03:39.080 --> 1:03:41.480
And the other one is artificial intelligence.
1:03:41.480 --> 1:03:45.120
Which promises to in some ways
1:03:45.120 --> 1:03:48.040
to do some of the programming for you
1:03:48.040 --> 1:03:49.440
is one way to think about it.
1:03:49.440 --> 1:03:52.280
So it's really what a programmer is,
1:03:52.280 --> 1:03:55.680
is using the intelligence that's inside your skull
1:03:55.680 --> 1:03:58.040
to do something useful.
1:03:58.040 --> 1:03:59.920
The hope with artificial intelligence is that
1:03:59.920 --> 1:04:02.040
it does some of the useful parts for you
1:04:02.040 --> 1:04:03.560
the way you don't have to think about it.
1:04:03.560 --> 1:04:06.000
So do you see smartphones,
1:04:06.000 --> 1:04:07.680
the fact that everybody has one
1:04:07.680 --> 1:04:09.320
and they're getting more and more powerful
1:04:09.320 --> 1:04:11.600
as potentially changing programming?
1:04:11.600 --> 1:04:14.280
And do you see AI as potentially changing programming?
1:04:14.280 --> 1:04:15.600
Oh, okay, so that's good.
1:04:15.600 --> 1:04:17.520
So smartphones have definitely changed.
1:04:17.520 --> 1:04:19.360
I mean, since, you know, I guess 2010
1:04:19.360 --> 1:04:21.800
is when they really started getting super popular.
1:04:21.800 --> 1:04:24.120
I mean, in the last eight years,
1:04:24.120 --> 1:04:25.680
the world has literally changed, right?
1:04:25.680 --> 1:04:28.480
Like everybody carries a computer around and that's normal.
1:04:28.480 --> 1:04:31.280
I mean, that is such a huge change in society.
1:04:31.280 --> 1:04:32.120
I think we're still dealing
1:04:32.120 --> 1:04:34.960
with a lot of the positive negative ramifications of that,
1:04:34.960 --> 1:04:35.800
right?
1:04:35.800 --> 1:04:36.620
Like everybody's connected all the time.
1:04:36.620 --> 1:04:37.800
Everybody's on the computer all the time.
1:04:37.800 --> 1:04:40.360
That was my dream world as a geek, right?
1:04:40.360 --> 1:04:42.040
But it's like, be careful what you ask for, right?
1:04:42.040 --> 1:04:44.120
Like, wow, now everybody has a computer
1:04:44.120 --> 1:04:46.320
and it's not quite the utopia that we thought it would be,
1:04:46.320 --> 1:04:47.160
right?
1:04:47.160 --> 1:04:48.520
Computers can be used for a lot of stuff
1:04:48.520 --> 1:04:51.200
that's not necessarily great.
1:04:51.200 --> 1:04:53.160
So to me, that's the central focus of the smartphone
1:04:53.160 --> 1:04:55.720
is just that it puts a computer in front of everyone,
1:04:55.720 --> 1:04:57.200
granted a small touch screen,
1:04:57.200 --> 1:04:59.120
smallish touch screen computer.
1:04:59.120 --> 1:05:00.560
But as for programming, like, I don't know,
1:05:00.560 --> 1:05:03.800
I don't think that I've kind of over time come to subscribe
1:05:03.800 --> 1:05:06.120
to the UNIX view of the world when it comes to programming.
1:05:06.120 --> 1:05:10.000
It's like, you wanna teach these basic command line things
1:05:10.000 --> 1:05:12.160
and that is just what programming is gonna be for,
1:05:12.160 --> 1:05:14.400
I think a long, long time.
1:05:14.400 --> 1:05:17.080
I don't think there's any magical, like visual programming
1:05:17.080 --> 1:05:18.160
that's gonna happen.
1:05:19.480 --> 1:05:20.600
I just, I don't know.
1:05:20.600 --> 1:05:22.720
I've over time have become a believer
1:05:22.720 --> 1:05:24.280
in that UNIX philosophy of just, you know,
1:05:24.280 --> 1:05:26.400
they kind of had a right with UNIX.
1:05:26.400 --> 1:05:28.920
That's gonna be the way it is for a long, long time.
1:05:28.920 --> 1:05:31.560
And we'll continue to, like I said, raise the baseline.
1:05:31.560 --> 1:05:33.000
The tools will get better, it'll get simpler,
1:05:33.000 --> 1:05:35.600
but it's still fundamentally gonna be command line tools,
1:05:35.600 --> 1:05:37.160
you know, fancy IDEs.
1:05:37.160 --> 1:05:39.200
That's kind of it for the foreseeable future.
1:05:39.200 --> 1:05:42.240
I'm not seeing any visual programming stuff on the horizon.
1:05:42.240 --> 1:05:43.080
Cause you kind of think like,
1:05:43.080 --> 1:05:44.080
what do you do on a smartphone
1:05:44.080 --> 1:05:46.400
that will be directly analogous to programming?
1:05:46.400 --> 1:05:47.480
Like I'm trying to think, right?
1:05:47.480 --> 1:05:49.600
Like, and there's really not much.
1:05:52.160 --> 1:05:55.160
So not necessarily analogous to programming,
1:05:55.160 --> 1:06:00.160
but the kind of things that,
1:06:00.840 --> 1:06:02.680
the kind of programs you need to write
1:06:02.680 --> 1:06:06.360
might need to be very different.
1:06:07.360 --> 1:06:08.200
Yeah.
1:06:08.200 --> 1:06:09.760
And the kind of languages, I mean,
1:06:09.760 --> 1:06:12.400
but I probably also subscribed to the same,
1:06:12.400 --> 1:06:14.280
just because everything in this world
1:06:14.280 --> 1:06:16.520
might be written in JavaScript.
1:06:16.520 --> 1:06:17.760
Oh yeah, that's definitely, that's already happening.
1:06:17.760 --> 1:06:19.000
I mean, discourse is a bet on,
1:06:19.000 --> 1:06:20.840
discourse is itself, JavaScript is another bet
1:06:20.840 --> 1:06:21.680
on that side of the table.
1:06:21.680 --> 1:06:23.680
And I still try and believe in that.
1:06:23.680 --> 1:06:25.840
So I would say smartphones have mostly a cultural shift,
1:06:25.840 --> 1:06:28.000
more than a programming shift.
1:06:28.000 --> 1:06:30.080
Now your other question was about artificial intelligence
1:06:30.080 --> 1:06:32.480
and like, sort of advice is predicting
1:06:32.480 --> 1:06:33.320
what you're gonna do.
1:06:33.320 --> 1:06:34.640
And I do think there's some strength to that.
1:06:34.640 --> 1:06:36.040
I think artificial intelligence
1:06:36.040 --> 1:06:37.680
kind of overselling it in terms of what it's doing.
1:06:37.680 --> 1:06:39.240
It's more like, people are predictable, right?
1:06:39.240 --> 1:06:40.760
People do the same things.
1:06:40.760 --> 1:06:42.240
Like, let me give you an example.
1:06:42.240 --> 1:06:44.600
One check we put into a discourse
1:06:44.600 --> 1:06:48.280
that's in a lot of big commercial websites is,
1:06:48.280 --> 1:06:51.160
say you log in from New York City now,
1:06:51.160 --> 1:06:54.240
and then an hour later, you log in from San Francisco.
1:06:54.240 --> 1:06:56.600
It's like, well, hmm, that's interesting.
1:06:56.600 --> 1:06:59.840
How did you get from New York to San Francisco in one hour?
1:06:59.840 --> 1:07:01.040
So at that point, you're like, okay,
1:07:01.040 --> 1:07:02.640
this is a suspicious login at that point.
1:07:02.640 --> 1:07:03.480
So we would alert you.
1:07:03.480 --> 1:07:05.360
It's like, okay, but that's not AI, right?
1:07:05.360 --> 1:07:07.280
That's just heuristic of like,
1:07:07.280 --> 1:07:11.080
how did you in one hour get 2,000 miles, right?
1:07:11.080 --> 1:07:12.680
That doesn't mean, maybe you're on a VPN,
1:07:12.680 --> 1:07:13.520
there's other ways to happen.
1:07:13.520 --> 1:07:16.000
But that's just a basic prediction based on the idea
1:07:16.000 --> 1:07:19.280
that people pretty much don't move around that much.
1:07:19.280 --> 1:07:20.920
Like they may travel occasionally,
1:07:20.920 --> 1:07:22.960
but like nobody, I mean, unless you're a traveling salesman
1:07:22.960 --> 1:07:25.520
that's literally traveling the world every day,
1:07:25.520 --> 1:07:28.240
like there's so much repetition and predictability
1:07:28.240 --> 1:07:29.600
in terms of things you're going to do.
1:07:29.600 --> 1:07:31.880
And I think good software anticipates your needs.
1:07:31.880 --> 1:07:34.160
Like for example, Google, I think it's called Google Now
1:07:34.160 --> 1:07:36.160
or whatever that Google thing is that predicts your commute
1:07:36.160 --> 1:07:37.720
and predicts based on your phone location,
1:07:37.720 --> 1:07:39.160
like where are you every day?
1:07:39.160 --> 1:07:42.080
Well, that's probably where you work, that kind of stuff.
1:07:42.080 --> 1:07:43.760
I do think computers can get a lot better at that,
1:07:43.760 --> 1:07:46.280
but I hesitate to call it like full blown AI.
1:07:46.280 --> 1:07:48.480
It's just computers getting better at like,
1:07:48.480 --> 1:07:49.480
first of all, they have a ton of data
1:07:49.480 --> 1:07:50.520
because everybody has a smartphone.
1:07:50.520 --> 1:07:52.000
Now, all of a sudden we have all this data
1:07:52.000 --> 1:07:53.840
that we didn't have before about location,
1:07:53.840 --> 1:07:56.160
about like, you know, communication
1:07:56.160 --> 1:07:59.200
and feeding that into some basic heuristics
1:07:59.200 --> 1:08:00.480
and maybe some fancy algorithms
1:08:00.480 --> 1:08:03.120
that turn it into predictions of anticipating your needs
1:08:03.120 --> 1:08:04.520
like a friend would, right?
1:08:04.520 --> 1:08:06.560
Like, oh, hey, I see your home,
1:08:06.560 --> 1:08:07.520
would you like some dinner, right?
1:08:07.520 --> 1:08:08.640
Like, let's go get some food
1:08:08.640 --> 1:08:10.760
because that's usually what we do this time of day, right?
1:08:10.760 --> 1:08:13.640
In the context of actually the active programming,
1:08:13.640 --> 1:08:15.360
do you see IDEs improving
1:08:15.360 --> 1:08:17.480
and making the life of programming is better?
1:08:17.480 --> 1:08:18.840
I do think that is possible
1:08:18.840 --> 1:08:20.720
because there's a lot of repetition in programming, right?
1:08:20.720 --> 1:08:23.200
Oh, you know, Clippy would be the bad example of,
1:08:23.200 --> 1:08:25.960
oh, I see, it looks like you're writing a for loop,
1:08:25.960 --> 1:08:27.840
but there are patterns in code, right?
1:08:27.840 --> 1:08:30.360
Like, and actually libraries are kind of like that, right?
1:08:30.360 --> 1:08:33.200
Like, rather than go, you know,
1:08:33.200 --> 1:08:35.440
code up your own HTTP request library,
1:08:35.440 --> 1:08:37.560
it's like, well, you'd use one of the existing ones
1:08:37.560 --> 1:08:39.720
that we have that's already a trouble shot, right?
1:08:39.720 --> 1:08:41.400
Like, it's not AI per se,
1:08:41.400 --> 1:08:44.800
it's just, you know, building better Lego bricks,
1:08:44.800 --> 1:08:47.720
bigger Lego bricks that have more functionality in them
1:08:47.720 --> 1:08:48.880
so people don't have to worry
1:08:48.880 --> 1:08:50.520
about the low level stuff as much anymore.
1:08:50.520 --> 1:08:52.880
Like WordPress, for example, to me is like,
1:08:52.880 --> 1:08:55.880
a tool for somebody who isn't a programmer to do something,
1:08:55.880 --> 1:08:57.640
I mean, you can turn WordPress into anything.
1:08:57.640 --> 1:08:59.480
It's kind of crazy actually through plugins, right?
1:08:59.480 --> 1:09:01.320
And that's not programming per se,
1:09:01.320 --> 1:09:04.520
it's just Lego bricks stacking WordPress elements, right?
1:09:04.520 --> 1:09:06.560
And a little bit of configuration glue.
1:09:06.560 --> 1:09:08.480
So I would say maybe in a broader sense,
1:09:08.480 --> 1:09:11.320
what I'm seeing like, there'll be more gluing
1:09:11.320 --> 1:09:14.120
and less like actual programming.
1:09:14.120 --> 1:09:15.560
And that's a good thing, right?
1:09:15.560 --> 1:09:17.920
Cause most of the stuff you need is kind of out there already.
1:09:17.920 --> 1:09:20.040
You said 1970s, Unix,
1:09:20.040 --> 1:09:30.680
do you see PHP and these kind of old remnants of, of the early birth
1:09:30.680 --> 1:09:33.720
of programming remaining with us for a long time?
1:09:33.720 --> 1:09:37.320
Like you said, Unix in itself, do you see ultimately,
1:09:37.320 --> 1:09:42.840
you know, this stuff's just being there out of momentum?
1:09:42.840 --> 1:09:44.080
I kind of do.
1:09:44.080 --> 1:09:46.120
I mean, I was a big believer in Windows early on
1:09:46.120 --> 1:09:47.320
and I was a big, you know, I was like,
1:09:47.320 --> 1:09:48.720
Unix, what a waste of time.
1:09:48.720 --> 1:09:50.080
But over time, I've completely flipped on that
1:09:50.080 --> 1:09:51.680
where I was like, okay, the Unix guys were right
1:09:51.680 --> 1:09:54.600
and pretty much Microsoft and Windows were kind of wrong,
1:09:54.600 --> 1:09:55.720
at least on the server side.
1:09:55.720 --> 1:09:57.400
And on the desktop, right, you need a GUI,
1:09:57.400 --> 1:09:58.240
you need a lot of stuff.
1:09:58.240 --> 1:09:59.160
And you have the two philosophies,
1:09:59.160 --> 1:10:02.640
like Apple built on Unix, effectively Darwin.
1:10:02.640 --> 1:10:04.560
And on the desktop, it's a slightly different story,
1:10:04.560 --> 1:10:06.960
but on the server side where you're going to be programming.
1:10:06.960 --> 1:10:08.440
Now it's a question of where the programming is going to be.
1:10:08.440 --> 1:10:10.720
There's going to be a lot more like client side programming
1:10:10.720 --> 1:10:13.600
cause technically discourse is client side programming.
1:10:13.600 --> 1:10:14.480
The way you get discourse,
1:10:14.480 --> 1:10:16.160
we deliver a big ball of JavaScript,
1:10:16.160 --> 1:10:18.240
which is then executed locally.
1:10:18.240 --> 1:10:20.960
So we're really using a lot more local computing power.
1:10:20.960 --> 1:10:21.960
We'll still retrieve the data.
1:10:21.960 --> 1:10:23.360
Obviously, we have to display the posts
1:10:23.360 --> 1:10:24.280
on the screen and so forth,
1:10:24.280 --> 1:10:27.000
but in terms of like sorting and a lot of the basic stuff,
1:10:27.000 --> 1:10:29.400
we're using the host processor.
1:10:29.400 --> 1:10:31.240
But to the extent that a lot of programming
1:10:31.240 --> 1:10:33.000
is still going to be server side,
1:10:33.000 --> 1:10:35.280
I would say, yeah, the Unix philosophy definitely won.
1:10:35.280 --> 1:10:38.640
And there'll be different veneers over the Unix,
1:10:38.640 --> 1:10:40.800
but it's still, if you peel away one or two layers,
1:10:40.800 --> 1:10:44.320
it's going to be Unix for a long, I think, Unix one,
1:10:44.320 --> 1:10:45.520
I mean, so definitively.
1:10:45.520 --> 1:10:47.440
It's interesting to hear you say that
1:10:47.440 --> 1:10:49.080
because you've done so much excellent work
1:10:49.080 --> 1:10:52.840
on the Microsoft side in terms of backend development.
1:10:52.840 --> 1:10:53.680
Cool.
1:10:53.680 --> 1:10:56.960
So what's the future hold for Jeff Atwood?
1:10:56.960 --> 1:10:59.120
I mean, the discourse,
1:10:59.120 --> 1:11:02.560
continuing the discourse in trying to improve
1:11:02.560 --> 1:11:03.840
conversation on the web?
1:11:03.840 --> 1:11:05.640
Well, discourse is what I believe is a,
1:11:05.640 --> 1:11:07.040
and originally I called a five year project,
1:11:07.040 --> 1:11:08.800
but then really quickly revised that to a 10 year project.
1:11:08.800 --> 1:11:12.280
So we started in early 2013,
1:11:12.280 --> 1:11:13.640
that's when we launched the first version.
1:11:13.640 --> 1:11:16.120
So we're still, you know, five years in.
1:11:16.120 --> 1:11:17.400
This is the part where it starts getting good,
1:11:17.400 --> 1:11:19.640
like we have a good product on discourse.
1:11:19.640 --> 1:11:21.800
There's any project you built in software,
1:11:21.800 --> 1:11:24.120
it takes three years to build what you wanted to build anyway.
1:11:24.120 --> 1:11:26.600
Like V1 is going to be terrible, which it was,
1:11:26.600 --> 1:11:27.440
but you ship it anyway,
1:11:27.440 --> 1:11:28.920
because that's how you get better at stuff.
1:11:28.920 --> 1:11:29.960
It's about turning the crank.
1:11:29.960 --> 1:11:31.440
It's not about V1 being perfect,
1:11:31.440 --> 1:11:32.840
because that's ridiculous.
1:11:32.840 --> 1:11:35.600
It's about V1, then let's get really good at V1.1,
1:11:35.600 --> 1:11:38.040
1.2, 1.3, like how fast can we iterate?
1:11:38.040 --> 1:11:40.160
And I think we're iterating like crazy on discourse,
1:11:40.160 --> 1:11:41.960
the point that like it's a really good product now,
1:11:41.960 --> 1:11:43.620
we have serious momentum.
1:11:43.620 --> 1:11:46.300
And my original vision was,
1:11:46.300 --> 1:11:48.180
I want to be the WordPress of discussion,
1:11:48.180 --> 1:11:49.900
meaning if someone came to you and said,
1:11:49.900 --> 1:11:50.820
I want to start a blog,
1:11:50.820 --> 1:11:53.020
although the very question is kind of archaic now,
1:11:53.020 --> 1:11:55.340
it's like who actually blogs anymore.
1:11:55.340 --> 1:11:59.060
But I wanted the answer to that to be,
1:11:59.060 --> 1:12:01.420
it would be WordPress normally,
1:12:01.420 --> 1:12:04.220
because that's the obvious choice for blogging most of the time.
1:12:04.220 --> 1:12:05.980
But if someone said, hey, I want to,
1:12:05.980 --> 1:12:08.780
I need a group of people to get together and do something,
1:12:08.780 --> 1:12:10.460
the answer should be discourse, right?
1:12:10.460 --> 1:12:11.820
That should be the default answer for people.
1:12:11.820 --> 1:12:13.420
Cause it's open source, it's free,
1:12:13.420 --> 1:12:16.020
doesn't cost you anything, you control it, you can run it.
1:12:16.020 --> 1:12:17.420
Your minimum server cost for discourse
1:12:17.420 --> 1:12:19.620
is five bucks a month at this point.
1:12:19.620 --> 1:12:21.580
They actually got the VPS prices down,
1:12:21.580 --> 1:12:23.900
it used to be $10 a month for one gigabyte of RAM,
1:12:23.900 --> 1:12:26.700
which we are dependent,
1:12:26.700 --> 1:12:28.460
we have a kind of heavy stack,
1:12:28.460 --> 1:12:30.340
like there's a lot of stuff in discourse.
1:12:30.340 --> 1:12:31.820
You need Postgres, you need Redis,
1:12:31.820 --> 1:12:33.900
you need Ruby on Rails,
1:12:33.900 --> 1:12:35.800
you need a sidekick for scheduling.
1:12:35.800 --> 1:12:36.980
It's not a trivial amount of stuff,
1:12:36.980 --> 1:12:38.060
cause we were architected for like,
1:12:38.060 --> 1:12:39.500
look, we're building for the next 10 years.
1:12:39.500 --> 1:12:41.980
I don't care about shared PHP hosting,
1:12:41.980 --> 1:12:44.180
that's not my model.
1:12:44.180 --> 1:12:45.340
My idea is like, hey, you know,
1:12:45.340 --> 1:12:47.540
eventually this is going to be very cheap for everybody.
1:12:47.540 --> 1:12:49.140
And I want to build it right,
1:12:49.140 --> 1:12:50.580
using again, you know,
1:12:50.580 --> 1:12:53.500
higher, bigger building block levels, right?
1:12:53.500 --> 1:12:54.460
That have more requirements.
1:12:54.460 --> 1:12:56.620
And there's a WordPress model of WordPress.org,
1:12:56.620 --> 1:12:57.700
WordPress.com.
1:12:57.700 --> 1:13:01.020
Is there a central hosting for discourse or no?
1:13:01.020 --> 1:13:02.740
There is, we're not strictly segmenting
1:13:02.740 --> 1:13:05.140
into the open source versus the commercial side.
1:13:05.140 --> 1:13:06.060
We have a hosting business,
1:13:06.060 --> 1:13:07.140
that's how discourse makes money,
1:13:07.140 --> 1:13:08.900
as we host discourse instances,
1:13:08.900 --> 1:13:09.980
and we have really close relationship
1:13:09.980 --> 1:13:12.860
with our customers of the symbiosis
1:13:12.860 --> 1:13:14.700
of them giving us feedback on the product.
1:13:14.700 --> 1:13:16.500
We definitely weight feedback from customers,
1:13:16.500 --> 1:13:18.420
a lot heavier than feedback from somebody
1:13:18.420 --> 1:13:20.940
who just wanders by and gives feedback.
1:13:20.940 --> 1:13:22.500
But that's where we make all our money,
1:13:22.500 --> 1:13:24.660
but we don't have a strict division.
1:13:24.660 --> 1:13:26.660
We encourage people to use discourse,
1:13:26.660 --> 1:13:29.060
like the whole point is that it's free, right?
1:13:29.060 --> 1:13:29.980
Anybody can set it up.
1:13:29.980 --> 1:13:32.620
I don't want to be the only person that hosts discourse,
1:13:32.620 --> 1:13:34.220
that's absolutely not the goal,
1:13:34.220 --> 1:13:36.300
but it is a primary way for us to build a business.
1:13:36.300 --> 1:13:37.620
And it's actually kind of a great business.
1:13:37.620 --> 1:13:40.300
I mean, the business is going really, really well
1:13:40.300 --> 1:13:41.380
in terms of hosting.
1:13:41.380 --> 1:13:44.340
So I used to work at Google Research,
1:13:44.340 --> 1:13:47.060
as a company that's basically funded on advertisements,
1:13:47.060 --> 1:13:50.420
so it's Facebook, let me ask if you can comment on it.
1:13:50.420 --> 1:13:53.540
I think advertisement at its best.
1:13:53.540 --> 1:13:56.620
So you'd be extremely critical on what ads are,
1:13:56.620 --> 1:13:59.740
but at its best, it's actually serving you,
1:13:59.740 --> 1:14:01.220
in a sense, it's giving you,
1:14:01.220 --> 1:14:05.700
it's connecting you to what you would want to explore.
1:14:05.700 --> 1:14:08.060
So it's like related posts or related content,
1:14:08.060 --> 1:14:10.220
it's the same, that's the best of advertisement.
1:14:10.220 --> 1:14:15.220
So discourse is connecting people based on their interests.
1:14:16.100 --> 1:14:20.100
It seems like a place where advertisement at its best
1:14:20.100 --> 1:14:21.780
could actually serve the users.
1:14:21.780 --> 1:14:24.940
Is that something that you're considering thinking about
1:14:24.940 --> 1:14:29.460
as a way to bring, to financially support the platform?
1:14:29.460 --> 1:14:31.380
That's interesting because I actually have
1:14:31.380 --> 1:14:32.620
a contrarian view of advertising,
1:14:32.620 --> 1:14:33.820
which I kind of agree with you.
1:14:33.820 --> 1:14:36.820
I recently installed Ad Blocker reluctantly
1:14:36.820 --> 1:14:38.300
because I don't like to do that,
1:14:38.300 --> 1:14:40.620
but the performance of the ads, man,
1:14:40.620 --> 1:14:43.340
they're so heavy now and it's just crazy.
1:14:43.340 --> 1:14:45.100
So it's almost like a performance argument
1:14:45.100 --> 1:14:47.220
more than I actually am pro ads,
1:14:47.220 --> 1:14:50.380
and I have a contrarian view point, I agree with you.
1:14:50.380 --> 1:14:51.980
If you do ads right, it's serving you stuff
1:14:51.980 --> 1:14:53.300
you would be interested in anyway.
1:14:53.300 --> 1:14:56.540
I don't mind that, that actually is kind of a good thing.
1:14:56.540 --> 1:15:00.020
So plus, I think it's rational to want to support
1:15:00.020 --> 1:15:01.700
the people that are doing this work,
1:15:01.700 --> 1:15:02.620
through seeing their ads,
1:15:02.620 --> 1:15:04.060
but that said, I run Ad Block now,
1:15:04.060 --> 1:15:06.220
which I didn't want to do,
1:15:06.220 --> 1:15:08.220
but I was convinced by all these articles,
1:15:08.220 --> 1:15:11.620
like 30, 40 megabytes of stuff just to serve you ads.
1:15:12.500 --> 1:15:16.020
Yeah, it feels like ads now are like the experts exchange
1:15:16.020 --> 1:15:18.260
of whenever you start to stack overflows.
1:15:18.260 --> 1:15:19.540
It's a little bit, it's overwhelming.
1:15:19.540 --> 1:15:21.060
Oh, there's so many companies in Adtech
1:15:21.060 --> 1:15:22.340
that what's embarrassing, like you can do that.
1:15:22.340 --> 1:15:24.300
Have you seen those logo charts of like just the whole page?
1:15:24.300 --> 1:15:26.260
It's like, you can't even see them, they're so small.
1:15:26.260 --> 1:15:27.820
There's so many companies in the space,
1:15:27.820 --> 1:15:29.660
but since you brought it up, I do want to point out
1:15:29.660 --> 1:15:31.900
that very, very few Discord sites actually run
1:15:31.900 --> 1:15:33.100
using an ad supported model.
1:15:33.100 --> 1:15:34.700
It's not effective.
1:15:34.700 --> 1:15:37.620
Like it's too diluted, it's too weird,
1:15:37.620 --> 1:15:40.540
it doesn't pay well, and like users hate it.
1:15:40.540 --> 1:15:42.820
So it's a combination of like users hate it,
1:15:42.820 --> 1:15:44.460
it doesn't actually work that well in practice.
1:15:44.460 --> 1:15:46.260
Like in theory, yes, I agree with you.
1:15:46.260 --> 1:15:48.740
Clean, fast ads that were exactly the stuff
1:15:48.740 --> 1:15:50.300
you would be interested in, awesome.
1:15:50.300 --> 1:15:52.460
We're so far from that though, right?
1:15:52.460 --> 1:15:53.780
Like, and Google does an okay job,
1:15:53.780 --> 1:15:55.380
they do retargeting and stuff like that,
1:15:55.380 --> 1:15:58.060
but in the real world,
1:15:58.060 --> 1:16:01.020
Discord sites rarely can make ads work.
1:16:01.020 --> 1:16:03.060
It just doesn't work for so many reasons.
1:16:03.060 --> 1:16:05.620
But you know what does work is subscriptions,
1:16:05.620 --> 1:16:10.620
Patreon, affiliate codes for like Amazon,
1:16:10.700 --> 1:16:13.140
of like just, oh, here's a cool YoYo click,
1:16:13.140 --> 1:16:14.420
and then you click and go to Amazon,
1:16:14.420 --> 1:16:15.700
they get a small percentage of that,
1:16:15.700 --> 1:16:17.180
which is fair, I think.
1:16:17.180 --> 1:16:19.140
I mean, because you saw the YoYo on that site,
1:16:19.140 --> 1:16:21.140
and you click through and you bought it, right?
1:16:21.140 --> 1:16:22.580
That's fair for them to get 5% of that
1:16:22.580 --> 1:16:24.220
or 2% of that, whatever it is.
1:16:24.220 --> 1:16:26.020
Those things definitely work.
1:16:26.020 --> 1:16:28.620
In fact, a site that I used to participate on a lot,
1:16:28.620 --> 1:16:29.900
I helped the owner.
1:16:29.900 --> 1:16:32.460
One of the things, I got them switched to Discord,
1:16:32.460 --> 1:16:34.220
I basically paid them to switch to Discord,
1:16:34.220 --> 1:16:35.380
because I was like, look, you guys gotta switch,
1:16:35.380 --> 1:16:38.540
I can't come here anymore on this terrible software.
1:16:38.540 --> 1:16:40.180
But I was like, look, and on top of that,
1:16:40.180 --> 1:16:42.300
like you're serving people ads that they hate,
1:16:42.300 --> 1:16:43.980
like you should just go full on Patreon,
1:16:43.980 --> 1:16:45.340
because he had a little bit of Patreon,
1:16:45.340 --> 1:16:49.740
go full on Patreon, do the Amazon affiliates thing
1:16:49.740 --> 1:16:51.140
for any Amazon links that get posted,
1:16:51.140 --> 1:16:53.700
and just do that, and just triple down on that stuff.
1:16:53.700 --> 1:16:55.340
And that's worked really well for them,
1:16:55.340 --> 1:16:56.820
and this creator in particular.
1:16:56.820 --> 1:16:59.500
So that stuff works, but traditional ads,
1:16:59.500 --> 1:17:01.700
I mean, definitely not working, at least on Discord.
1:17:01.700 --> 1:17:05.940
So last question, you've created the code keyboard.
1:17:05.940 --> 1:17:07.820
I've programmed most of my adult life
1:17:07.820 --> 1:17:09.860
in a Kinesis keyboard.
1:17:09.860 --> 1:17:12.100
I have one upstairs now.
1:17:12.100 --> 1:17:14.540
Can you describe what a mechanical keyboard is,
1:17:14.540 --> 1:17:16.740
and why is it something that makes you happy?
1:17:16.740 --> 1:17:18.780
Well, you know, this is another fetish item, really.
1:17:18.780 --> 1:17:20.140
Like, it's not required.
1:17:20.140 --> 1:17:22.260
You can do programming on any kind of keyboard, right?
1:17:22.260 --> 1:17:24.420
Even like an onscreen keyboard, oh God,
1:17:24.420 --> 1:17:25.940
that's terrifying, right?
1:17:25.940 --> 1:17:28.500
But you could, I mean, if you look back at the early days
1:17:28.500 --> 1:17:29.780
of computing, there were chiclet keyboards,
1:17:29.780 --> 1:17:31.460
which are, I mean, those are awful, right?
1:17:31.460 --> 1:17:32.540
But what's a chiclet keyboard?
1:17:32.540 --> 1:17:34.180
Oh God, okay, well, it's just like
1:17:34.180 --> 1:17:36.220
thin rubber membranes.
1:17:36.220 --> 1:17:37.700
Oh, the rubber ones, oh no.
1:17:37.700 --> 1:17:38.540
Super bad, right?
1:17:38.540 --> 1:17:39.380
Yeah.
1:17:39.380 --> 1:17:40.380
So it's a fetish item.
1:17:40.380 --> 1:17:41.940
All that really says is, look,
1:17:41.940 --> 1:17:43.340
I care really about keyboards,
1:17:43.340 --> 1:17:44.620
because the keyboard is the primary method
1:17:44.620 --> 1:17:46.260
of communication with the computer, right?
1:17:46.260 --> 1:17:49.340
So it's just like having a nice mic for this podcast.
1:17:49.340 --> 1:17:50.620
You want a nice keyboard, right?
1:17:50.620 --> 1:17:52.180
Because it has very tactile feel.
1:17:52.180 --> 1:17:53.980
I can tell exactly when I press the key.
1:17:53.980 --> 1:17:56.900
I get that little click, so oh, and it feels good.
1:17:56.900 --> 1:17:57.980
And it's also kind of a fetish item.
1:17:57.980 --> 1:18:00.260
It's like, wow, I care enough about programming
1:18:00.260 --> 1:18:02.380
that I care about the tool, the primary tool,
1:18:02.380 --> 1:18:03.300
that I use to communicate with the computer
1:18:03.300 --> 1:18:06.380
and make sure it's as good as it feels good to use for me.
1:18:06.380 --> 1:18:08.300
And like, I can be very productive with it.
1:18:08.300 --> 1:18:10.700
So to be honest, it's a little bit of a fetish item,
1:18:10.700 --> 1:18:11.540
but a good one.
1:18:11.540 --> 1:18:12.540
It indicates that you're serious,
1:18:12.540 --> 1:18:13.580
it indicates you're interested.
1:18:13.580 --> 1:18:15.260
It indicates that you care about the fundamentals,
1:18:15.260 --> 1:18:17.500
because you know what makes you a good programmer?
1:18:17.500 --> 1:18:18.900
Being able to type really fast, right?
1:18:18.900 --> 1:18:20.580
Like, this is true, right?
1:18:20.580 --> 1:18:23.500
So a core skill is just being able to type fast enough
1:18:23.500 --> 1:18:26.140
to get your ideas out of your head into the code base.
1:18:26.140 --> 1:18:29.740
So just practicing your typing can make you a better programmer.
1:18:29.740 --> 1:18:33.740
It is also something that makes you,
1:18:33.740 --> 1:18:36.660
well, makes you enjoy typing, correct?
1:18:36.660 --> 1:18:39.940
The actual act, something about the process,
1:18:39.940 --> 1:18:41.220
like I play piano.
1:18:41.220 --> 1:18:42.060
It's tactile.
1:18:42.060 --> 1:18:46.260
There's a tactile feel that ultimately feeds the passion,
1:18:46.260 --> 1:18:47.100
makes you happy.
1:18:47.100 --> 1:18:48.220
Right, no, totally, that's it.
1:18:48.220 --> 1:18:49.060
I mean, and it's funny,
1:18:49.060 --> 1:18:50.700
because artisanal keyboards have exploded,
1:18:50.700 --> 1:18:53.460
like Mastrop has gone ballistic with this stuff.
1:18:53.460 --> 1:18:56.900
There's probably like 500 keyboard projects on Mastrop alone.
1:18:56.900 --> 1:18:58.180
And there's some other guy I follow on Twitter.
1:18:58.180 --> 1:19:00.300
I used to write for this, the site, the tech report,
1:19:00.300 --> 1:19:01.420
way back in the day.
1:19:01.420 --> 1:19:02.940
And he's like, every week he's just posting like,
1:19:02.940 --> 1:19:05.460
what I call keyboard porn of like, just cool keyboards.
1:19:05.460 --> 1:19:06.620
Oh my God, those look really cool, right?
1:19:06.620 --> 1:19:09.820
Like, that's like, how many keyboards this guy have, right?
1:19:09.820 --> 1:19:10.740
It's got like me with yoyos.
1:19:10.740 --> 1:19:11.580
How many yoyos do you have?
1:19:11.580 --> 1:19:12.420
How many do you need?
1:19:12.420 --> 1:19:14.900
Well, technically one, but I like a lot.
1:19:14.900 --> 1:19:15.820
I don't know why.
1:19:15.820 --> 1:19:17.260
So same thing with keyboards.
1:19:17.260 --> 1:19:18.940
So yeah, they're awesome.
1:19:18.940 --> 1:19:21.580
Like I highly recommend anybody who doesn't have a mechanical
1:19:21.580 --> 1:19:23.100
to research it, look into it,
1:19:23.100 --> 1:19:24.420
and see what you like.
1:19:24.420 --> 1:19:26.300
And you know, it's ultimately a fetish item,
1:19:26.300 --> 1:19:28.860
but I think these sort of items,
1:19:28.860 --> 1:19:31.020
these religious artifacts that we have
1:19:31.020 --> 1:19:32.220
are part of what make us human.
1:19:32.220 --> 1:19:33.940
Like that part's important, right?
1:19:33.940 --> 1:19:35.780
It's kind of what makes life worth living.
1:19:35.780 --> 1:19:38.900
Yeah, it's not necessary in the strictest sense,
1:19:38.900 --> 1:19:42.340
but ain't nothing necessary if you think about it, right?
1:19:42.340 --> 1:19:44.300
Like, so yeah, why not?
1:19:44.300 --> 1:19:45.420
So sure.
1:19:45.420 --> 1:19:47.500
Jeff, thank you so much for talking today.
1:19:47.500 --> 1:19:48.340
Yeah, you're welcome.
1:19:48.340 --> 1:19:53.340
Thanks for having me.